RU2513900C1 - Method and device to determine object coordinates - Google Patents

Abstract

FIELD: measurement equipment.

SUBSTANCE: method is realised due to more accurate measurement of a direction vector to an object $$\overrightarrow{V_{{П}_{ij}}}$$

in the system of coordinates of a video camera, reduction of accidental errors of assessment due to multiple detection of object coordinates by series of frames, and also due to taking into account features of area relief in the area of measurements.

EFFECT: increased accuracy of detection of object coordinates.

5 cl, 26 dwg

Description

The inventive objects are united by a single inventive concept, relate to the field of radio engineering, namely to passive location, and can be used in navigation, direction-finding, location-based tools for visual detection and determination of coordinates of a priori unknown objects from flight-lifting means (LPS).

Known methods for determining the coordinates of moving and stationary objects according to US Pat. RF 2251712 and US Pat. RF 2154284, providing the determination of the coordinates of objects using optoelectronic devices. They involve determining the angular coordinate of the image of the object along with image-changing elements in the field of view, followed by the conversion of the obtained value into a stabilized coordinate system, determining the magnitude and direction of the linear velocity of the object in a stabilized coordinate system, the formation of the angular displacement in the stabilized picture plane based on the obtained value and coordinates characterizing the linear displacement of the image-changing elements relative to their own coordinate system object, and adjusting the angular coordinate of the image of the object together with image-distorting elements in the stabilized coordinate system by the value of the angular displacement. However, these systems have a significant drawback. Analog methods are implemented on the earth's surface and have a small radius of action.

The known method according to US Pat. RF 2323851, IPC B64C 31/06, publ. 05/10/2008. An analogue involves the use of an unmanned aerial vehicle (UAV) in conjunction with a video camera. An analogue provides photographing the earth's surface under the control of a ground post.

The analogue also has the disadvantage associated with the inability to determine the coordinates of the detected objects.

The closest in technical essence is a method for determining the location of a source of radio emission according to US Pat. RF №2465613, IPC G01S 3/14, publ. 10/27/2012

, $${\text{L}}_{\text{lps}}{}_i$$

, $${\text{H}}_{\text{lps}}{}_i$$

, соответственно широта, долгота и высота ЛПС, и его пространственную ориентацию (k_k, l_k, ζ_k)_i, где $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

соответственно углы крена, тангажа и склонения ЛПС в i-й момент времени, совместное запоминание навигационных и временных параметров ЛПС, а при обнаружении j-го заданного источника радиоизлучения (ИРИ) в момент времени t_i предварительное определение его пространственно-информационных параметров: азимута θ_ij и угла места β_ij в системе координат антенной системы, перевод координат ЛПС в геоцентрическую систему координат, а вектора направления на j-й объект $$\overrightarrow{V_{{}_{ij}}}={\left(\theta ,{\beta }_j\right)}_i$$

в левостороннюю систему декартовых координат $$\overrightarrow{V_{{П}_{ij}}}={(X_j,Y_j,Z_j)}_i$$

, коррекцию вектора направления на j-й ИРИ $$\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}}}={(X_{{}_j}^{\text{'}},Y_{{}_j}^{\text{'}},Z_{{}_j}^{\text{'}})}_i$$

с учетом априорно известной ориентации антенной системы пеленгатора относительно борта ЛПС (k_ant, l_ant, ζ_ant) путем последовательного умножения значений $$\overrightarrow{V_{{П}_{ij}}}$$

на соответствующие углам Эйлера матрицы поворота, после чего в нормальной системе координат выполнение вычисления уточненного значения вектора направления $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

на j-й ИРИ с учетом измеренных в момент времени t_i пространственных углов ЛПС: крена $${\text{k}}_{\text{lps}}{}_i$$

, тангажа $${\text{l}}_{\text{lps}}{}_i$$

и склонения $${\zeta }_{\text{lps}}{}_i$$

, определение уточненных значений азимута $${\theta }_{ij}^{+}$$

, угла места $${\beta }_{ij}^{+}$$

и удаления ЛПС, находящегося в момент времени t_i на высоте $${\text{H}}_{\text{lps}}{}_i$$

, от j-го ИРИ d(Н₀)_ij, расположенного на поверхности "круглой" Земли, определение в геоцентрической системе координат значения истинного вектора направления на j-й объект $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

, которое зависит от широты B_lps, долготы L_lps местоположения ЛПС, определение координаты точки пересечения вектора $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

с "круглой" Землей $$\overrightarrow{V_{фj}}$$

, преобразование геоцентрических координат j-го объекта $$\overrightarrow{V_{фj}}$$

в географические $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, где В_j и L_j соответственно широта и долгота местоположения j-го ИРИ, определение удаления j-го ИРИ относительно координат ЛПС по параллели dL_ij, меридиану dB_ij и перпендикуляру (высоте) dH_ij, вычисление предварительных значений азимутального угла $${\theta }_{kj}^{\text{'}}$$

, угла места $${\beta }_{kj}^{\text{'}}$$

настройки видеокамеры без учета пространственной ориентации ЛПС и видеокамеры, преобразование сферических координат $${\theta }_{kj}^{\text{'}}$$

и $${\beta }_{kj}^{\text{'}}$$

j-го ИРИ в нормальную систему координат $$\overrightarrow{V_{2_j}}={(X_2,Y_2,Z_2)}_j$$

и далее в систему координат видеокамеры $$\overrightarrow{V_{1_j}}={(X_1,Y_1,Z_1)}_j$$

с учетом пространственной ориентации ЛПС и видеокамеры, определение истинных значений азимутального угла θ_kj и угла места β_kj ориентации видеокамеры на i-й ИРИ, одновременное оценивание угла закрытия корпусом ЛПС направления на j-й ИРИ, а при выполнении условия β_kj<0 ориентирование видеокамеры в соответствии с параметрами θ_kj и β_kj.The preparatory method at the preparatory stage includes the installation of a video camera under the LPS fuselage, determining the orientation of the video camera and antenna system (AS) of the direction finder relative to the LPS side (k_k, l_k, ζ_k), (k_antl_antζ_ant) where k_k, l_k, ζ_k, k_antl_antζ_ant - respectively, the roll, pitch and declination angles of the video camera and the speaker, and during the flight, the location of the LPS is constant after a specified time interval Δt (B_lps, L_lps, H_lps)_i, Where $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

, $${\text{H}}_{\text{lps}}{}_i$$

, respectively, the latitude, longitude and height of the LPS, and its spatial orientation (k_k, l_k, ζ_k)_iwhere $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

 accordingly, the angles of roll, pitch and declination of the LPS at the i-th moment of time, the joint storage of navigation and time parameters of the LPS, and upon detection of the j-th given source of radio emission (IRI) at time t_i preliminary determination of its spatial information parameters: azimuth θ_ij and elevation angle β_ij in the coordinate system of the antenna system, translation of the LPS coordinates into a geocentric coordinate system, and the direction vector to the j-th object $$\overrightarrow{V_{{}_{ij}}}={\left(\theta ,{\beta }_j\right)}_i$$

 to the left-handed Cartesian coordinate system $$\overrightarrow{V_{P_{ij}}}={(X_j,Y_j,Z_j)}_i$$

, direction vector correction on the j-th IRI $$\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}\text{'}}}={(X_{{}_j}^{\text{'}\text{'}},Y_{{}_j}^{\text{'}\text{'}},Z_{{}_j}^{\text{'}\text{'}})}_i$$

 taking into account the a priori known orientation of the direction finder antenna system relative to the LPS side (k_antl_antζ_ant) by sequentially multiplying the values $$\overrightarrow{V_{P_{ij}}}$$

 to the rotation matrix corresponding to the Euler angles, after which, in a normal coordinate system, the calculation of the updated value of the direction vector $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

 on the j-th IRI taking into account measured at time t_i spatial angles of LPS: roll $${\text{k}}_{\text{lps}}{}_i$$

pitch $${\text{l}}_{\text{lps}}{}_i$$

 and declensions $${\zeta }_{\text{lps}}{}_i$$

, determination of refined azimuth values $${\theta }_{ij}^{+}$$

elevation $${\beta }_{ij}^{+}$$

 and removing the LPS located at time t_i on high $${\text{H}}_{\text{lps}}{}_i$$

, from the jth IRI d (Н₀)_ijlocated on the surface of the "round" Earth, determination in the geocentric coordinate system of the value of the true direction vector to the j-th object $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

which depends on latitude B_lps, longitudes L_lps location of the LPS, determining the coordinates of the intersection points of the vector $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

 with round earth $$\overrightarrow{V_{fj}}$$

, transformation of geocentric coordinates of the j-th object $$\overrightarrow{V_{fj}}$$

 into geographical $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, where in_j and L_j respectively, the latitude and longitude of the location of the j-th IRI, determining the removal of the j-th IRI relative to the LPS coordinates along the parallel dL_ij, meridian dB_ij and perpendicular (height) dH_ij, calculation of preliminary values of the azimuthal angle $${\theta }_{kj}^{\text{'}\text{'}}$$

elevation $${\beta }_{kj}^{\text{'}\text{'}}$$

 video camera settings without taking into account the spatial orientation of LPS and video cameras, conversion of spherical coordinates $${\theta }_{kj}^{\text{'}\text{'}}$$

 and $${\beta }_{kj}^{\text{'}\text{'}}$$

 j-Iran into the normal coordinate system $$\overrightarrow{V_{2_j}}={({\mathrm{<figure-callout\; id="2"\; label="X"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">X</figure-callout>}}_2,{\mathrm{<figure-callout\; id="2"\; label="Y"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">Y</figure-callout>}}_2,Z_2)}_j$$

 and further into the coordinate system of the camcorder $$\overrightarrow{V_{{\mathrm{one}}_j}}={(X_{\mathrm{one}},Y_{\mathrm{one}},Z_{\mathrm{one}})}_j$$

 taking into account the spatial orientation of the LPS and video camera, determining the true values of the azimuthal angle θ_kj and elevation angle β_kj orientation of the camcorder to the i-th IRI, simultaneous estimation of the angle of closure by the LPS housing of the direction to the j-th IRI, and when condition β is satisfied_kj<0 camera orientation in accordance with the parameters θ_kjand β_kj.

The method allows to determine the location of the IRI using a video camera and radio direction finder from radio emissions. The use of a video camera in the prototype improves the accuracy of the meter.

The prototype method also has a disadvantage associated with the relatively low accuracy of determining the coordinates of objects. In practice, as a rule, the interest is not the source of radio emission itself, but the object on which (or in which) it is located. In a situation when the IRI at the facility is off, the prototype loses its functionality. In addition, Iran is often taken out of the facility, which also exacerbates the situation. The low accuracy of the measurements is explained by the fact that a rather large area of the earth's surface is observed in the video camera, much exceeding the area of the object, which entails errors in determining its coordinates (see figa).

The purpose of the proposed technical solution is to increase the accuracy of determining the coordinates of the object from the LPS.

, $${\text{L}}_{\text{lps}}{}_i$$

, $${\text{H}}_{\text{lps}}{}_i$$

, соответственно широта, долгота и высота ЛПС, и его пространственную ориентацию (k_lps, l_lps, ζ_lps), где $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

соответственно углы крена, тангажа и склонения ЛПС в i-й момент времени, совместно запоминают навигационные и временные параметры ЛПС, а при визуальном обнаружении j-го заданного объекта в момент времени t_i предварительно определяют вектор направления на него $$\overrightarrow{V_{{П}_{ij}}}={(X_j,Y_j,Z_j)}_i$$

в системе координат видеокамеры, переводят координаты ЛПС в геоцентрическую систему координат, корректируют вектор направления на j-й объект $$\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}}}={(X_{{}_j}^{\text{'}},Y_{{}_j}^{\text{'}},Z_{{}_j}^{\text{'}})}_i$$

с учетом априорно известной ориентации камеры относительно борта ЛПС (k_k, l_k, ζ_k) путем последовательного умножения значений $$\overrightarrow{V_{{П}_{ij}}}$$

на соответствующие углам Эйлера матрицы поворота, после чего в нормальной системе координат вычисляют уточненное значение вектора направления $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

на j-й объект с учетом измеренных в момент времени t_i пространственных углов ЛПС: крена $${\text{k}}_{\text{lps}}{}_i$$

, тангажа $${\text{l}}_{\text{lps}}{}_i$$

и склонения $${\zeta }_{\text{lps}}{}_i$$

, определяют уточненные значения азимута $${\theta }_{ij}^{+}$$

, угла места $${\beta }_{ij}^{+}$$

и удаление ЛПС, находящегося в момент времени t_i на высоте $${\text{H}}_{\text{lps}}{}_i$$

, от j-го объекта d(H₀)_ij, расположенного на поверхности "круглой" Земли, в геоцентрической системе координат определяют значение истинного вектора направления на j-й объект $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

, которое зависит от широты B_lps, долготы L_lps местоположения ЛПС, определяют координаты точки пересечения вектора $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

с "круглой" Землей $$\overrightarrow{V_{фj}}$$

, преобразуют геоцентрические координаты j-го объекта $$\overrightarrow{V_{фj}}$$

в географические $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, где В_j и L_j соответственно широта и долгота местоположения j-го объекта, отличающийся тем, что предварительно измеряют и запоминают коэффициенты дисторсии объектива видеокамеры k₁, k₂ и k₃, положение видеокамеры относительно борта ЛПС фиксируют на весь период измерений, в качестве j-го объекта может выступать любой стационарный или подвижный физический объект, наблюдаемый в видеокамеру, причем решение о необходимости измерения координат наблюдаемого объекта принимает оператор, значение предварительного вектора направления на j-й объект $$\overrightarrow{V_{{П}_{ij}}}={(X_j,Y_j,Z_j)}_i$$

определяют по местоположению объекта на кадре в момент времени t_i, уточняют значение предварительного вектора направления на j-й объект $$\overrightarrow{V_{{П}_{ij}}}$$

путем устранения влияния на результаты измерений дисторсии объектива видеокамеры, при наличии n последовательных кадров, n=2, 3, …, N, с изображением j-го объекта выполняют n циклов измерений географических координат $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, а результаты измерений усредняют, при наличии цифровой карты местности района измерений, представляющей собой матрицу с заданной дискретностью по координатам района измерений с соответствующими значениями высот рельефа, уточняют географические координаты обнаруженного j-го объекта.This goal is achieved by the fact that in the known method for determining the coordinates of objects, which consists in the fact that at the preparatory stage, a video camera is installed on board the flight-lifting means (LPS) under the fuselage, the orientation of the video camera relative to the LPS side is determined (k_k, l_k, ζ_k), where k_k, l_k, ζ_k - respectively, the angles of roll, pitch and declination of the camera, and during the flight, the location of the LPS is constantly determined after a specified time interval Δt (B_lps, L_lps, H_lps), where $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

, $${\text{H}}_{\text{lps}}{}_i$$

, respectively, the latitude, longitude and height of the LPS, and its spatial orientation (k_lps, l_lps, ζ_lps), where $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

 accordingly, the angles of roll, pitch and declination of the LPS at the i-th moment in time, together remember the navigation and time parameters of the LPS, and when the j-th target is visually detected at time t_i pre-determine the direction vector to it $$\overrightarrow{V_{P_{ij}}}={(X_j,Y_j,Z_j)}_i$$

 in the coordinate system of the video camera, translate the LPS coordinates into a geocentric coordinate system, adjust the direction vector to the j-th object $$\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}\text{'}}}={(X_{{}_j}^{\text{'}\text{'}},Y_{{}_j}^{\text{'}\text{'}},Z_{{}_j}^{\text{'}\text{'}})}_i$$

 taking into account the a priori known orientation of the camera relative to the LPS side (k_k, l_k, ζ_k) by sequentially multiplying the values $$\overrightarrow{V_{P_{ij}}}$$

 to the rotation matrix corresponding to the Euler angles, after which, in the normal coordinate system, the updated value of the direction vector is calculated $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

 to the j-th object, taking into account measured at time t_i spatial angles of LPS: roll $${\text{k}}_{\text{lps}}{}_i$$

pitch $${\text{l}}_{\text{lps}}{}_i$$

 and declensions $${\zeta }_{\text{lps}}{}_i$$

, determine the specified azimuth values $${\theta }_{ij}^{+}$$

elevation $${\beta }_{ij}^{+}$$

 and removing the LPS at time t_i on high $${\text{H}}_{\text{lps}}{}_i$$

, from the jth object d (H₀)_ijlocated on the surface of the "round" Earth, in the geocentric coordinate system determine the value of the true direction vector to the j-th object $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

which depends on latitude B_lps, longitudes L_lps LPS locations, determine the coordinates of the intersection point of the vector $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

 with round earth $$\overrightarrow{V_{fj}}$$

transform geocentric coordinates of the j-th object $$\overrightarrow{V_{fj}}$$

 into geographical $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, where in_j and L_j respectively, the latitude and longitude of the location of the j-th object, characterized in that the distortion coefficients of the camera lens k are previously measured and stored_one, k₂ and k₃, the position of the video camera relative to the LPS board is fixed for the entire measurement period, any stationary or moving physical object observed in the video camera can act as the j-th object, and the operator decides whether to measure the coordinates of the observed object, the value of the preliminary direction vector to the j-th an object $$\overrightarrow{V_{P_{ij}}}={(X_j,Y_j,Z_j)}_i$$

 determined by the location of the object on the frame at time t_i, specify the value of the preliminary direction vector to the j-th object $$\overrightarrow{V_{P_{ij}}}$$

 by eliminating the effect on the distortion results of the camera lens, in the presence of n consecutive frames, n = 2, 3, ..., N, with the image of the j-th object, n cycles of measurements of geographical coordinates are performed $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

and the measurement results are averaged, in the presence of a digital map of the terrain of the measurement region, which is a matrix with a given discreteness in the coordinates of the measurement region with the corresponding elevation heights, the geographical coordinates of the detected j-th object are specified.

The coordinates of the j-th object (x _r , y _r ) _j in the frame are determined in pixels, counted from the upper left corner of the frame of the video camera.

в системе координат видеокамеры осуществляют в соответствии с выражениемThe transition from the coordinates of the j-th object in the frame to the refined preliminary vector of direction to it $$\overrightarrow{V_{P_{ij}}}$$

in the coordinate system of the camera is carried out in accordance with the expression

, $$\overrightarrow{V_{P_{ij}}}={\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}/\psi \left(\overrightarrow{V_{P_{ij}}}\right)$$

,

, $${\tilde{x}}_{{}_r}={\left(x_r,y_r\right)}^T$$

, $$\tilde{c}={\left(c_x,c_y\right)}^T$$

- координаты центра матрицы (кадра) в пикселях, f - фокусное расстояние объектива видеокамеры, пересчитанное в пиксели матрицы, $$\psi \left(\overrightarrow{V_{{П}_{ij}}}\right)=\left(1+k_1{\Vert \overrightarrow{V_{{П}_{ij}}}\Vert }^2+k_2{\Vert \overrightarrow{V_{{П}_{ij}}}\Vert }^4+k_3{\Vert \overrightarrow{V_{{П}_{ij}}}\Vert }^6\right)$$

, k₁, k₂, k₃ - измеренные коэффициенты дисторсии объектива, с использованием метода простых итераций.Where $${\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}=\left({\tilde{x}}_r-\tilde{c}\right)/f$$

, $${\tilde{x}}_{{}_r}={\left(x_r,y_r\right)}^T$$

, $$\tilde{c}={\left(c_x,c_y\right)}^T$$

- coordinates of the center of the matrix (frame) in pixels, f is the focal length of the lens of the camera, converted into pixels of the matrix, $$\psi \left(\overrightarrow{V_{P_{ij}}}\right)=\left(\mathrm{one}+k_{\mathrm{one}}{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^2+{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">k</figure-callout>}}_2{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^{\mathrm{four}}+k_3{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^6\right)$$

, k ₁ , k ₂ , k ₃ are the measured lens distortion coefficients using the simple iteration method.

на j-й объект, определяется заданной точностью предварительного измерения координат объекта, рассчитывают координаты $$\overrightarrow{V_{ф,j,m}}$$

, соответствующие дискретно выделенным высотам рельефа местности Н_i,m, а за предварительные координаты j-го объекта $$\overrightarrow{V_{ф,j,m}}$$

принимают первую точку разбиения вектора $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

, находящуюся ниже уровня рельефа местности, уточняют местоположение j-го объекта путем выделения соседней точки разбиения $$\overrightarrow{V_{ф,j,m-1}}$$

, находящейся над рельефом местности, отрезок $$\left(\overrightarrow{V_{ф,j,m-1}},\overrightarrow{V_{ф,j,m}}\right)$$

вектора направления на j-й объект $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

делят на δ равных интервалов, Δδ<<Δd, где Δδ - шаг сканирования по выделенному отрезку вектора направления $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}}}$$

, и определяется конечной заданной точностью измерения координат объектов, для названных точек вычисляют координаты $$\overrightarrow{V_{ф,j,\delta }}$$

и соответствующие им значения высоты рельефа местности H_i,m,δ, за точные координаты j-го объекта принимают значение $$\overrightarrow{V_{ф,j,m,\delta }}$$

, находящееся между соседними точками, расположенными выше и ниже рельефа местности, а полученное значение координат j-го объекта $$\overrightarrow{V_{ф,j,m,\delta }}$$

преобразуют в удобную географическую систему координат $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

.In the presence of a digital terrain map of the terrain of the measurement region, a sequential set of elevation values {H _{i, m} }, m = 1, 2, ..., M, which corresponds to uniformly distributed coordinates on the segment connecting the coordinates (B _lps , L _lps ) and (B _j , L _j ), M = d (H ₀ ) / Δd, where Δd is the scanning step along the direction vector $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

on the j-th object, determined by the specified accuracy of the preliminary measurement of the coordinates of the object, calculate the coordinates $$\overrightarrow{V_{f,j,m}}$$

corresponding to discrete selected heights of the terrain Н _{i, m} , and for the preliminary coordinates of the j-th object $$\overrightarrow{V_{f,j,m}}$$

take the first split point of the vector $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

, located below the level of the terrain, specify the location of the j-th object by highlighting the neighboring break point $$\overrightarrow{V_{f,j,m-\mathrm{one}}}$$

located above the terrain $$\left(\overrightarrow{V_{f,j,m-\mathrm{one}}},\overrightarrow{V_{f,j,m}}\right)$$

direction vector to j-th object $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

divided into δ equal intervals, Δδ << Δd, where Δδ is the scanning step along the selected segment of the direction vector $$\overrightarrow{V_{{}_j}^{\text{'}\text{'}\text{'}\text{'}}}$$

, and is determined by the final specified accuracy of measuring the coordinates of objects, for the named points, coordinates are calculated $$\overrightarrow{V_{f,j,\delta }}$$

and the corresponding values of the height of the terrain H _{i, m, δ} , for the exact coordinates of the j-th object take the value $$\overrightarrow{V_{f,j,m,\delta }}$$

located between neighboring points located above and below the terrain, and the obtained value of the coordinates of the j-th object $$\overrightarrow{V_{f,j,m,\delta }}$$

convert to a convenient geographic coordinate system $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

.

направления на объект благодаря учету его местоположения на кадре видеокамеры с возможностью последующего усреднения результатов оценивания по серии кадров, а итоговые усредненные координаты объекта уточняют в соответствии с особенностями рельефа местности района измерений.Thanks to the new set of features in the claimed method, a more accurate measurement of the preliminary vector is achieved. $$\overrightarrow{V_{P_{ij}}}$$

directions to the object due to taking into account its location on the frame of the video camera with the possibility of subsequent averaging of the estimation results over a series of frames, and the final averaged coordinates of the object are specified in accordance with the terrain features of the measurement area.

Known devices according to US Pat. RF №2251712, IPC G01S 13/66; US Pat. RF №2359288, IPC G01S 5/02. Analogs provide the determination of the coordinates of objects using optical-electronic means. However, they have a significant drawback: they are located on the earth's surface, as a result of which they have a small radius of action.

The closest in its technical essence to the claimed device for determining the coordinates of objects is pat. RF №2323851, IPC B64C 31/06 "Earth observation system with unmanned aerial vehicle", publ. 05/10/2008

The prototype device contains a tightening winch, unmanned aerial vehicle (UAV) and ground control station (NPU), and the UAV consists of a series-connected controller, steering gear and aerodynamic rudders, an autopilot, the group of information inputs of which are connected to the second group of information outputs of the controller, the first a group of information inputs of which is connected to a group of information outputs of an autopilot, a propulsion system, a group of information inputs of which is connected to a third group of inf controller outputs, a video surveillance unit, the group of information inputs of which is connected to the fourth group of information outputs of the controller, and the group of information outputs - with the second group of information inputs of the controller, and the first transceiver module, the group of information inputs of which is connected to the fifth group of information outputs of the controller, the third group the information inputs of which are connected to the group of information outputs of the first transceiver module, and the ground control point is nen comprising a series-connected control unit, a second receiving-transmitting unit and a processing and displaying device, the second group control unit outputs a control bus NPU, and is connected to the clamping winch.

The prototype is designed to take photographs of the earth's surface from a height of about 100 meters. However, the prototype device does not provide measurement of the coordinates of visible objects due to the lack of a high-precision UAV navigation system and a high-precision system for determining the direction to it.

The purpose of the proposed technical solution is to develop a device that provides high-precision measurement of the coordinates of specified objects from the UAV.

This goal is achieved by the fact that in the known device, consisting of an unmanned aerial vehicle and ground control station, and the UAV is made containing serially connected controller, steering gear and aerodynamic steering wheels, autopilot, the group of information inputs of which are connected to the second group of information outputs of the controller, the first group the information inputs of which are connected to the group of information outputs of the autopilot, a propulsion system, the group of information inputs of which is connected with the third group of information outputs of the controller, the first transceiver module, the group of information inputs of which is connected to the fourth group of information outputs of the controller, the second group of information inputs of which is connected to the group of information outputs of the first transceiver module, and the video surveillance unit, and the ground control station is made containing connected by the first control unit, the second transceiver module and the first information processing and display device, I distinguish in that the transmitting module, the UAV navigation unit and the storage device are additionally introduced into the UAV, moreover, the first group of information inputs of the storage device is connected to the group of information outputs of the video surveillance unit, the second group of information inputs is connected to the group of information outputs of the UAV navigation unit, and the group of information of the outputs of the storage device is connected to a group of information inputs of the transmitting module, and a serial connection is additionally introduced to the ground control point nennye receiving unit and the second processing unit and display unit, the second control unit, the group of information inputs of which is combined with the first group of information inputs of the second processing device and display information, and the group of information outputs - to the second group of information inputs of the second processing device and display information.

The above-mentioned new set of essential features due to the fact that new elements and connections are being found allows achieving the objective of the invention: to provide high-precision measurement of the coordinates of the given objects due to the full and objective measurement of the spatial parameters of the UAV, the camera’s orientation angles and the object’s position in the frame under the influence of destabilizing factors (wind load, UAV maneuvers, etc.) and taking into account the terrain features.

The inventive objects are illustrated by drawings, which show:

figure 1 illustrates:

a) a spot of illumination of a video camera on the earth’s surface with an object О _j located in it;

b) the cause of errors in determining the coordinates of the object in the absence of taking into account the terrain;

figure 2 shows a generalized algorithm for determining the coordinates of the object;

figure 3 shows a generalized structural diagram of a device for determining the coordinates of an object;

figure 4 shows the algorithm for determining the coordinates of the object (B, L, H) _j in the geographical coordinate system in one frame;

figure 5 illustrates the algorithm for finding the coordinates of the object [B, L, H) _j in the geographical coordinate system for a series of frames;

figure 6 illustrates the shooting of a flat surface from two different positions of the camcorder;

7 illustrates the procedure for preliminary determination of the coordinates of the object;

on Fig explains the procedure for determining the coordinates of the object with a given accuracy;

figure 9 shows the algorithm of the video surveillance subsystem placed on board the UAV;

figure 10 shows the algorithm of the second automated workstation of the ground control point;

11 illustrates a structural diagram of a second device for processing and displaying information;

on Fig shows a structural diagram of a seventh calculator;

on Fig shows a structural diagram of a block distortion correction lens of the camera;

Fig. 14 is a structural diagram of an image processing unit;

on Fig illustrates the algorithm of the seventh calculator to determine the direction vector to the object in the coordinate system of the camera;

on Fig shows the algorithm of the processing unit of the video image;

on Fig shows the algorithm of the eighth calculator;

on Fig shows the algorithm of the first calculator;

Fig.19 shows the algorithm of the second calculator;

on Fig illustrates the algorithm of the third computer;

in Fig.21 shows the algorithm of the fourth calculator;

in Fig.22 shows the algorithm of the fifth transmitter;

on Fig shows the algorithm of the sixth calculator;

on Fig shows the algorithm of the ninth transmitter;

on Fig the algorithm of the third control unit in the mode of preliminary determination of coordinates;

Fig.26 shows the algorithm of the second control unit.

, тангажа $${\text{l}}_{\text{lps}}{}_i$$

и курсового угла $${\alpha }_{\text{lps}}{}_i$$

(угла сноса или склонения $${\zeta }_{\text{lps}}{}_i$$

), где $${\zeta }_{\text{lps}}{}_i={\mu }_{lp{s}_i}-{\alpha }_{lp{s}_i}$$

, $${\mu }_{lp{s}_i}$$

- значение путевого угла в момент времени t_i. Информацию об этих параметрах можно получить с помощью известных устройств (см. пат. РФ №2374659, МПК G01IS 7/00, опубл. 27.11.2009 г. бюл. 33; пат. РФ №2371733, МПК G01S 5/10, опубл. 27.10.2009 г. бюл. 30). Поэтому основной акцент в описании предлагаемого изобретения сделан на особенностях преобразования изображения объекта на кадре видеокамеры в вектор направления $$\overrightarrow{V_{{П}_{ij}}}={(X_j,Y_j,Z_j)}_i$$

на него в системе координат видеокамеры с целью повышения точностных характеристик измерителя.Modern methods for determining the coordinates of objects with LPS, which implement the goniometric-range-finding method of positioning, are usually based on the use of global navigation satellite systems (GNSS) (see V.S.Shebshaevich, P.P.Dmitriev, N.V. Ivantsevich, etc. Satellite radio navigation systems / Under the editorship of V.S.Shebshaevich. - M.: Radio and communications, 1993, pp. 262-275). However, in addition to information about the LPS own location and the direction of its movement when solving location problems, it is necessary to know its spatial orientation: $${\text{k}}_{\text{lps}}{}_i$$

pitch $${\text{l}}_{\text{lps}}{}_i$$

and heading angle $${\alpha }_{\text{lps}}{}_i$$

(drift or declination angle $${\zeta }_{\text{lps}}{}_i$$

), where $${\zeta }_{\text{lps}}{}_i={\mu }_{lp{s}_i}-{\alpha }_{lp{s}_i}$$

, $${\mu }_{lp{s}_i}$$

- the value of the track angle at time t _i . Information on these parameters can be obtained using known devices (see US Pat. RF No. 2374659, IPC G01IS 7/00, publ. 11/27/2009 bull. 33; US Pat. RF No. 2371733, IPC G01S 5/10, publ. October 27, 2009 bull. 30). Therefore, the main emphasis in the description of the invention is made on the features of the conversion of the image of the object on the frame of the camera into the direction vector $$\overrightarrow{V_{P_{ij}}}={(X_j,Y_j,Z_j)}_i$$

on it in the coordinate system of the camera in order to improve the accuracy characteristics of the meter.

, тангажа $${\text{l}}_{\text{lps}}{}_i$$

и склонения $${\zeta }_{\text{lps}}{}_i$$

(k_k, l_k, ζ_k) относительно корпуса ЛПС (см. фиг.2, 4). Полученные значения запоминают и в последующем используют для уточнения результатов измерений. Кроме того, измеряются и запоминаются коэффициенты дисторсии k₁, k₂, k₃ объектива видеокамеры.The implementation of the proposed method is illustrated as follows. At the preparatory stage, a video camera is installed under the LPS fuselage and its orientation is recorded. The orientation of the video camera is measured in three planes adopted in aviation as a roll $${\text{k}}_{\text{lps}}{}_i$$

pitch $${\text{l}}_{\text{lps}}{}_i$$

and declensions $${\zeta }_{\text{lps}}{}_i$$

(k _k , l _k , ζ _k ) relative to the housing of the LPS (see figure 2, 4). The obtained values are stored and subsequently used to refine the measurement results. In addition, distortion coefficients k ₁ , k ₂ , k _{3 of} the camera lens are measured and stored.

During the flight of the LPS along a certain route, a specified camera is searched for with a video camera. When the j-th object is visually detected, the operator makes a decision on the need to measure its coordinates. To this end, the location of the object on the frame of the camcorder is initially set (determined) by the operator using the sight.

на кадре (см фиг.1а) задаются в пикселях, при этом их отсчитывают от верхнего левого угла кадра.The obtained coordinates of the j-th object $${\left(x_r,y_r\right)}_{{}^j}^T$$

on the frame (see figa) are set in pixels, while they are counted from the upper left corner of the frame.

в кадре на направление на объект в системе координат видеокамеры $$\overrightarrow{V_{{П}_{ij}}}$$

. Данная операция выполняется в соответствии с выражениемThe next step is to transform the coordinates of the j-th object $${\left(x_r,y_r\right)}_{{}^j}^T$$

in the frame to the direction of the object in the coordinate system of the camera $$\overrightarrow{V_{P_{ij}}}$$

. This operation is performed in accordance with the expression

$$\left(\begin{array}{c}{x}_r\\ y_r\end{array}\right)=f\cdot \left(\begin{array}{c}{x}^{\text{'}\text{'}\text{'}\text{'}}\\ y^{\text{'}\text{'}\text{'}\text{'}}\end{array}\right)+\left(\begin{array}{c}{c}_x\\ c_y\end{array}\right),\left(\mathrm{one}\right)$$

,Where $$\{\begin{array}{c}{x}^{\text{'}\text{'}\text{'}\text{'}}=x^{\text{'}\text{'}}\left(\mathrm{one}+k_{\mathrm{one}}{\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">r</figure-callout>}}^2+k_2{r}^{\mathrm{four}}+k_3{r}^6\right)\\ y^{\text{'}\text{'}\text{'}\text{'}}=y^{\text{'}\text{'}}\left(\mathrm{one}+k_{\mathrm{one}}{\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">r</figure-callout>}}^2+k_2{r}^{\mathrm{four}}+k_3{r}^6\right)\end{array}$$

,

, $$r=\sqrt{x^{\text{'}\text{'}2}+y^{\text{'}\text{'}2}}$$

,

выражение 1 принимает видf is the focal length of the lens of the camera, converted into pixels of the matrix (frame), (s _x , s _y ) ^T are the coordinates of the center of the matrix in pixels, k ₁ , k ₂ , k ₃ are the measured distortion coefficients of the lens (see Szeliski, Richard. Computer: Algorithms and Applications. Sprintger, 2010). By reassigning $${\left(x,y\right)}^T=\tilde{x}$$

expression 1 takes the form

$${\tilde{x}}_r=f\cdot {\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}+\tilde{c},\left(2\right)$$

. Вектор $${\tilde{x}}^{\text{'}\text{'}}$$

определяется на основе известных параметров видеокамеры и координат j-го объекта на кадреWhere $${\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}=V_{P_{ij}}\left(\mathrm{one}+k_{\mathrm{one}}{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^2+{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">k</figure-callout>}}_2{\Vert V\overrightarrow{{}_{P_{ij}}}\Vert }^{\mathrm{four}}+k_3{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^6\right)$$

. Vector $${\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}$$

determined based on the known parameters of the camera and the coordinates of the j-th object on the frame

$${\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}=\left({\tilde{x}}_r-\tilde{c}\right)/f.\left(3\right)$$

, получаем уравнение, решением которого является искомый вектор $$\overrightarrow{V_{{}_{ij}}}$$

Thus, denoting $$\psi (\tilde{x})=\left(\mathrm{one}+k_{\mathrm{one}}{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^2+{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">k</figure-callout>}}_2{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^{\mathrm{four}}+k_3{\Vert \overrightarrow{V_{P_{ij}}}\Vert }^6\right)$$

, we obtain an equation whose solution is the desired vector $$\overrightarrow{V_{{}_{ij}}}$$

$$\overrightarrow{V_{P_{ij}}}={\tilde{x}}^{\text{'}\text{'}\text{'}\text{'}}/\psi \left(\overrightarrow{V_{P_{ij}}}\right).\left(\mathrm{four}\right)$$

находится методом простых итераций.The numerical value of the vector $$\overrightarrow{V_{P_{ij}}}$$

found by simple iteration.

определено без учета ориентации видеокамеры относительно корпуса ЛПС и собственно ориентации ЛПС в пространстве. В силу названных причин в предлагаемом способе на первом этапе учитывают ориентацию видеокамеры относительно ЛПС и положение ЛПС в пространстве. Это достигается путем последовательного перехода из одной системы координат в другую, что удобнее и быстрее выполнять в декартовой системе координат. На втором этапе определения координат объектов учитывают особенности рельефа местности района измерений. Данные этапы достаточно полно освещены в пат. РФ №2419106, МПК G01S 13/46, опубл. 20.05.2011 г. и пат. РФ №2458360 МПК G01S 13/46, 5/02, 3/14, опубл. 1.08.2012 г.Direction to j-th object $$\overrightarrow{V_{{}_{ij}}}$$

determined without taking into account the orientation of the camera relative to the LPS body and the actual orientation of the LPS in space. For these reasons, in the proposed method at the first stage take into account the orientation of the camera relative to the LPS and the position of the LPS in space. This is achieved by a sequential transition from one coordinate system to another, which is more convenient and faster to perform in a Cartesian coordinate system. At the second stage of determining the coordinates of objects, the terrain features of the measurement region are taken into account. These steps are fully covered in US Pat. RF №2419106, IPC G01S 13/46, publ. 05/20/2011 and Pat. RF №2458360 IPC G01S 13/46, 5/02, 3/14, publ. 01/08/2012

, измеренные в момент времени t_i, преобразуют в геоцентрическую систему координат:LPS coordinates $$\overrightarrow{V_{P_{ij}}}=\left(B_{lps},L_{lps},H_{lps}\right)$$

measured at time t _i transform into a geocentric coordinate system:

$$\overrightarrow{V_{lp{s}_i}^G}=\left(\begin{array}{c}{X}_{lp{s}_i}\\ Y_{lp{s}_i}\\ Z_{lp{s}_i}\end{array}\right)=\left(\begin{array}{c}\left(H_{lp{s}_i}+R\right)\cos \left(B_{lp{s}_i}\right)\cos \left(L_{lp{s}_i}\right)\\ (H_{lp{s}_i}+R)\cos (B_{lp{s}_i})\sin (L_{lp{s}_i})\\ (H_{lp{s}_i}+R)\sin (B_{lp{s}_i})\end{array}\right).\left(5\right)$$

учитывают априорно известную ориентацию видеокамеры относительно ЛПС на основе данных, полученных на подготовительном этапе. Коррекцию $$\overrightarrow{V_{{П}_{ij}}}$$

осуществляют в плоскости трех углов Эйлера: крена k_k, тангажа l_k и склонения ζ_k. Исходный вектор $$\overrightarrow{V_{{П}_{ij}}}$$

последовательно перемножают на три соответствующие углам Эйлера матрицы поворота (см. фиг.2)In the first transformation of the direction vector to the jth object $$\overrightarrow{V_{P_{ij}}}$$

take into account the a priori known orientation of the camera relative to the LPS based on data obtained at the preparatory stage. Correction $$\overrightarrow{V_{P_{ij}}}$$

carried out in the plane of three Euler angles: roll k _k , pitch l _k and declination ζ _k . Source vector $$\overrightarrow{V_{P_{ij}}}$$

successively multiplied by three rotation matrices corresponding to Euler angles (see FIG. 2)

$$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}=E_3({\zeta }_k)E_2(l_k)E_{\mathrm{one}}(k_k)\overrightarrow{V_{P_{ij}}},\left(6\right)$$

,Where $$E_{\mathrm{one}}(k_k)=\left[\begin{array}{ccc}\mathrm{one}& 0& 0\\ 0& \cos (k_k)& -\sin (k_k)\\ 0& \sin (k_k)& \cos (k_k)\end{array}\right]$$

,

, $$E_2(l_k)=\left[\begin{array}{ccc}\cos (l_k)& 0& \sin (l_k)\\ 0& \mathrm{one}& 0\\ -\sin (l_k)& 0& \cos (l_k)\end{array}\right]$$

,

. $$E_{\mathrm{one}}({\zeta }_k)=\left[\begin{array}{ccc}\cos ({\zeta }_k)& -\sin ({\zeta }_k)& 0\\ \sin ({\zeta }_k)& \cos ({\zeta }_k)& 0\\ 0& 0& \mathrm{one}\end{array}\right]$$

.

с целью учета ориентации ЛПС относительно земной поверхности и положения ЛПС в пространстве, что позволяет получить уточненное значение вектора направления на j-й объект $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}={(X_{{}_j}^{+},Y_{{}_j}^{+},Z_{{}_j}^{+})}_i$$

. Переход через эту систему координат продиктован тем, что в ней измеряются углы ориентации ЛПС. Получение вектора направления на источник $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

в нормальной системе координат также предпочтительно. Во-первых, вычисление уточненного угла места $${\beta }_{ij}^{+}$$

возможно только в рассматриваемой системе координат, так как это фактически угол отклонения направления на объект от горизонтальной плоскости в точке нахождения ЛПС. Во-вторых, в этой системе координат удобно решать задачу определения точки пересечения вектора направления на объект с "круглой" Землей в силу того обстоятельства, что одна из осей системы направлена к центру Земли. Ориентация ЛПС обычно задается углами $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, и $${\zeta }_{\text{lps}}{}_i$$

, которые определяют в каждой точке относительно плоскости, касательной к сферической модели земной поверхности. Ось крена $${\text{k}}_{\text{lps}}{}_i$$

лежит в этой плоскости и направлена на географический север, ось склонения $${\zeta }_{\text{lps}}{}_i$$

перпендикулярна указанной плоскости и направлена к центру земли, ось тангажа $$l_{\text{lps}}{}_i$$

лежит в указанной плоскости таким образом, что тройка осей представляет правую декартову систему координат (см. Авиация: Энциклопедия. - М.: Большая Российская энциклопедия, 1994 г.). Полученный на предыдущем этапе вектор $$\overrightarrow{V_{{}_{ij}}^{\text{'}}}$$

последовательно перемножают на три соответствующие матрицы поворота (относительно каждой из названных осей)At the next stage, in the normal coordinate system, the corrected direction vector is converted to the jth object $$\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}\text{'}}}={(X_{{}_j}^{\text{'}\text{'}},Y_{{}_j}^{\text{'}\text{'}},Z_{{}_j}^{\text{'}\text{'}})}_i$$

in order to take into account the orientation of the LPS relative to the earth's surface and the position of the LPS in space, which allows you to get the updated value of the direction vector to the j-th object $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}={(X_{{}_j}^{+},Y_{{}_j}^{+},Z_{{}_j}^{+})}_i$$

. The transition through this coordinate system is dictated by the fact that LPS orientation angles are measured in it. Getting the direction vector to the source $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

in a normal coordinate system is also preferable. First, the calculation of the adjusted elevation angle $${\beta }_{ij}^{+}$$

it is possible only in the coordinate system under consideration, since this is actually the angle of deviation of the direction of the object from the horizontal plane at the location of the LPS. Secondly, in this coordinate system it is convenient to solve the problem of determining the point of intersection of the direction vector of the object with the "round" Earth due to the fact that one of the axes of the system is directed toward the center of the Earth. LPS orientation is usually given by angles $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, and $${\zeta }_{\text{lps}}{}_i$$

which are determined at each point relative to a plane tangent to a spherical model of the earth's surface. Roll axis $${\text{k}}_{\text{lps}}{}_i$$

lies in this plane and is directed to geographical north, the declination axis $${\zeta }_{\text{lps}}{}_i$$

perpendicular to the indicated plane and directed towards the center of the earth, pitch axis $$l_{\text{lps}}{}_i$$

lies in the indicated plane in such a way that the three axes represent the right Cartesian coordinate system (see Aviation: Encyclopedia. - M.: Big Russian Encyclopedia, 1994). The vector obtained in the previous step $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

successively multiplied by three corresponding rotation matrices (relative to each of the named axes)

$$\overrightarrow{V_{{}_{ij}}^{+\text{'}\text{'}}}=E_3{({\zeta }_{lps})}_i{E}_2{(l_{lps})}_i{E}_{\mathrm{one}}{(k_{lps})}_i\overrightarrow{V_{{}_{{}_{ij}}}^{\text{'}\text{'}}},\left(7\right)$$

,Where $$E_{\mathrm{one}}{(k_{lps})}_i=\left[\begin{array}{ccc}\mathrm{one}& 0& 0\\ 0& \cos {(k_{lps})}_i& -\sin {(k_{lps})}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}\\ 0& \sin {(k_{lps})}_i& \cos {(k_{lps})}_i\end{array}\right]$$

,

, $$E_2{(l_{lps})}_i=\left[\begin{array}{ccc}\cos {(l_{lps})}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}& 0& \sin {(l_{lps})}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}\\ 0& \mathrm{one}& 0\\ -\sin {(l_{lps})}_i& 0& \cos {(l_{lps})}_i\end{array}\right]$$

,

$$E_3{({\zeta }_{lps})}_i=\left[\begin{array}{ccc}\cos {({\zeta }_{lps})}_i& -\sin {({\zeta }_{lps})}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}& 0\\ \sin {({\zeta }_{lps})}_i& \cos {({\zeta }_{lps})}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}& 0\\ 0& 0& \mathrm{one}\end{array}\right]$$

и угла места $${\beta }_{ij}^{+}$$

определяют из выражения (7) следующим образом:Refined azimuth values $${\theta }_{ij}^{+}$$

and elevation $${\beta }_{ij}^{+}$$

determined from the expression (7) as follows:

$${\beta }_{ij}^{+}=\arccos \left(\frac{Z_{ij}^{+}}{\Vert V_{ij}^{+}\Vert }\right)-\frac{\mathrm{<figure-callout\; id="2"\; label="\pi "\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">\pi </figure-callout>}}{2};\left(8\right)$$

находится в нормальной системе координат: ОХ⁺ - направление на север, OY⁺ - на восток, OZ⁺ - к центру Земли.Here is the vector $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

is in the normal coordinate system: OX ⁺ - direction to the north, OY ⁺ - to the east, OZ ⁺ - to the center of the Earth.

пересечется с "круглой" Землей на высоте d(H₀)_ij метров:To find the distance between the LPS and the jth object d (H ₀ ) _ij, it is necessary to take into account the spherical nature of the Earth's surface. Otherwise, this problem can be interpreted as finding the distance d (/ H ₀ ) _ij at which the vector $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

intersects with the “round” Earth at a height of d (H ₀ ) _ij meters:

$$d{(H_0)}_{ij}=(H_{lp{s}_i}+R)\sin {\beta }_{ij}^{+}-\sqrt{D},\left(10\right)$$

, R - радиус Земли, R=6370000 м. Следует отметить, что расстояние d (Н₀)_ij возможно определить при условии D≥0. В противном случае начинают новый цикл измерений пространственных параметров j-го объекта $${\theta }_{i+\mathrm{1,}j}^{+}$$

и $${\beta }_{i+\mathrm{1,}j}^{+}$$

.where D is the discriminant of the quadratic equation: $$D={({\mathrm{<figure-callout\; id="0"\; label="H"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">H</figure-callout>}}_0+R)}^2-{(H_{lp{s}_i}+R)}^2{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2{\beta }_{ij}^{+}$$

, R is the radius of the Earth, R = 6370000 m. It should be noted that the distance d (Н ₀ ) _ij can be determined under the condition D≥0. Otherwise, they begin a new cycle of measuring the spatial parameters of the j-th object $${\theta }_{i+\mathrm{one,}j}^{+}$$

and $${\beta }_{i+\mathrm{one,}j}^{+}$$

.

, расположена с некоторым поворотом, который зависит от широты и долготы местоположения ЛПС. Для окончательного перехода в геоцентрическую систему координат необходимо довернуть вектор $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

на широту ЛПС и π/2 минус долготу ЛПС $${\text{L}}_{\text{lps}}{}_i$$

, используя матрицы поворота, а затем перенести центр системы координат в центр Земли, используя геоцентрические координаты ЛПС. В результате имеем истинный вектор $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

направления на j-й объектThe normal coordinate system in which the specified vector is located at this stage $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

, is located with some rotation, which depends on the latitude and longitude of the location of the LPS. For the final transition to the geocentric coordinate system, you must rotate the vector $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

latitude LPS and π / 2 minus longitude LPS $${\text{L}}_{\text{lps}}{}_i$$

using rotation matrices, and then transfer the center of the coordinate system to the center of the Earth using the geocentric coordinates of the LPS. As a result, we have a true vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

directions to the j-th object

$$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}=G_2{(L_{lps})}_i{G}_{\mathrm{one}}{\left(-B_{lps}+\frac{\pi }{2}\right)}_i\cdot \overrightarrow{V_{ij}^{+}},\left(\mathrm{eleven}\right)$$

,Where $$G_{\mathrm{one}}{\left(-B_{lps}+\frac{\pi }{2}\right)}_i=\left[\begin{array}{ccc}\cos {\left(-B_{lps}+\frac{\pi }{2}\right)}_i& 0& -\sin {\left(-B_{lps}+\frac{\pi }{2}\right)}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}\\ 0& \mathrm{one}& 0\\ -\sin {\left(-B_{lps}+\frac{\pi }{2}\right)}_i& 0& \cos {\left(-B_{lps}+\frac{\pi }{2}\right)}_i\end{array}\right]$$

,

. $$G_2{\left(L_{lps}\right)}_i=\left[\begin{array}{ccc}\cos {\left(L_{lps}\right)}_i& -\sin {\left(L_{lps}\right)}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}& 0\\ \sin {\left(L_{lps}\right)}_i& \cos {\left(L_{lps}\right)}_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="00000251.png,00000255.png"\; state="\{\{state\}\}">i</figure-callout>}}& 0\\ 0& 0& \mathrm{one}\end{array}\right]$$

.

This completes the first phase of the measurements.

сравнивают с пороговым значением Δβ, определяющим заданную потенциальную точность измерения местоположения объектов. Следует отметить, что угол места на источник $${\beta }_{ij}^{+}=0$$

соответствует горизонту, $${\beta }_{ij}^{+}=-{90}^o$$

соответствует зениту.In the next step, the elevation angle calculation results $${\beta }_{ij}^{+}$$

compared with a threshold value Δβ, which determines a given potential accuracy of measuring the location of objects. It should be noted that the elevation angle to the source $${\beta }_{ij}^{+}=0$$

corresponds to the horizon $${\beta }_{ij}^{+}=-{90}^o$$

corresponds to the zenith.

The coordinates of the object on the "round" Earth in a geocentric coordinate system in direction and distance at a height of H ₀ can be found using the expression:

$$\overrightarrow{V_{fj}}=\left(\begin{array}{c}{X}_{fj}\\ Y_{fj}\\ Z_{fj}\end{array}\right)=\overrightarrow{V_{{}_{lp{s}_i}}^G}+d{(H_0)}_{ij}\overrightarrow{V_{ij}^{"}}.\left(12\right)$$

осуществляют следующим образом:Transition from (12) to a more convenient geographical coordinate system $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

carried out as follows:

latitude $$B_j=\arcsin \left(Z_{fj}/\Vert \overrightarrow{V_{fj}}\Vert \right),\left(13\right)$$

Where $$\Vert \overrightarrow{V_{fj}}\Vert =\sqrt{{\mathrm{<figure-callout\; id="2"\; label="X\; f\; j"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">X\; </figure-callout>}}_{fj}^{}+{\mathrm{<figure-callout\; id="2"\; label="Y\; f\; j"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">Y\; </figure-callout>}}_{fj}^{}+{\mathrm{<figure-callout\; id="2"\; label="Z\; f\; j"\; filenames="00000240.png,00000241.png"\; state="\{\{state\}\}">Z\; </figure-callout>}}_{fj}^{}};$$

.longitude

.

(обеспечивается низкая точность измерения координат объекта), а также при выполнении пороговых условий и отсутствии цифровой карты рельефа местности района измерений, определяют координаты точки пересечения истинного вектора направления на j-й объект $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

с "круглой" Землей, которые далее поступают на выход и используются в качестве искомой величины.For small values $${\beta }_{ij}^{+}$$

(low accuracy of measuring the coordinates of the object is ensured), as well as when the threshold conditions are met and there is no digital map of the terrain of the measurement region, the coordinates of the point of intersection of the true direction vector to the j-th object are determined $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

with the "round" Earth, which then go to the exit and are used as the desired value.

и наличия цифровой карты рельефа района измерений становится возможным более точное измерение координат, которое в свою очередь выполняют в два этапа. На первом этапе формируют последовательный набор значений высот {H_i,m}, m=1, 2, …, М, которые соответствуют равномерно распределенным координатам на отрезке, соединяющем координаты ЛПС ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) и j-го объекта (B_j, L_j) (см. фиг.7). При этом количество названных точек М находится из соотношения: М=d(Н₀)_ij/Δd, где Δd - шаг сканирования по вектору направления $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

на j-й объект в момент времени t_i. Значение Δd определяется заданной точностью определения координат объекта на первом (предварительном) этапе измерений, например, Δd=500 м. Рассчитывают координаты $$\overrightarrow{V_{фim}}$$

, соответствующие дискретно выделенным высотам рельефа местности Н_im. За предварительные координаты j-го объекта $$\overrightarrow{V_{фs_{i}j}}$$

принимают первую точку разбиения вектора $$\overrightarrow{V_j}$$

, находящуюся ниже уровня рельефа местности.When the threshold conditions are met $${\beta }_{ij}^{+}>\Delta \beta$$

 and the availability of a digital elevation map of the measurement region, it becomes possible to more accurately measure the coordinates, which in turn is performed in two stages. At the first stage, a sequential set of heights {H_{i, m}}, m = 1, 2, ..., M, which correspond to uniformly distributed coordinates on the segment connecting the LPS coordinates ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) and j-th object (B_j, L_j) (see Fig. 7). Moreover, the number of named points M is found from the relation: M = d (N₀)_ij/ Δd, where Δd is the scan step along the direction vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

 to the jth object at time t_i. The value Δd is determined by the specified accuracy of determining the coordinates of the object at the first (preliminary) measurement stage, for example, Δd = 500 m. The coordinates are calculated $$\overrightarrow{V_{fim}}$$

corresponding to discrete selected heights of the terrain N_im. For the preliminary coordinates of the j-th object $$\overrightarrow{V_{f{s}_{i}j}}$$

take the first split point of the vector $$\overrightarrow{V_j}$$

below ground level.

, находящейся над рельефом местности (см. фиг.8). Отрезок $$\left(\overrightarrow{V_{{ф}_{j,m-1}},}\overrightarrow{V_{{ф}_{j,m}}}\right)$$

вектора направления на j-й объект $$\overrightarrow{V_j}$$

делят на δ равных интервалов, Δδ<<Δδ, где Δδ - шаг сканирования по выделенному отрезку истинного вектора направления $$\overrightarrow{V_j}$$

. Последний определяется конечной заданной точностью измерения координат объектов и разрешающей способностью (дискретностью) цифровой карты местности. Для названных точек вычисляют координаты $$\overrightarrow{V_{{ф}_{j,m,\delta }}}$$

и соответствующие им значения высот рельефа местности H_j,m,δ. За точные координаты j-го объекта на основе линейной интерполяции принимают значение $$\overrightarrow{V_{{ф}_{j,m,\delta }}}$$

, находящееся между соседними точками p и p-1, p∈δ, расположенные выше и ниже рельефа местности At the second stage of measurements, the location of the j-th object is clarified by selecting the neighboring split point $$\overrightarrow{V_{f_{j,m-\mathrm{one}}}}$$

located above the terrain (see Fig.8). Line segment $$\left(\overrightarrow{V_{f_{j,m-\mathrm{one}}},}\overrightarrow{V_{f_{j,m}}}\right)$$

direction vector to j-th object $$\overrightarrow{V_j}$$

divided into δ equal intervals, Δδ << Δδ, where Δδ is the scanning step along the selected segment of the true direction vector $$\overrightarrow{V_j}$$

. The latter is determined by the final specified accuracy of measuring the coordinates of objects and the resolution (discreteness) of a digital terrain map. For the named points, the coordinates are calculated $$\overrightarrow{V_{f_{j,m,\delta }}}$$

and the corresponding elevation heights H _{j, m, δ} . For the exact coordinates of the j-th object based on linear interpolation, take the value $$\overrightarrow{V_{f_{j,m,\delta }}}$$

located between adjacent points p and p-1, p∈δ located above and below the terrain

$$\overrightarrow{V_{f_{j,m,\delta }}}=\overrightarrow{V_{f_{j,m,p}}}+\left(\overrightarrow{V_f{}_{j,m,p-\mathrm{one}}}-\overrightarrow{V_{f_{j,m,p}}}\right)\cdot \frac{a}{a+b},\left(\mathrm{fourteen}\right)$$

, $$b=\left|\Vert \overrightarrow{V_{{ф}_{j,m,p-1}}}\Vert -R-H_{j,p-1}\right|$$

.Where

, $$b=|\; |\; |\Vert \overrightarrow{V_{f_{j,m,p-\mathrm{one}}}}\Vert -R-H_{j,p-\mathrm{one}}|\; |\; |$$

.

преобразуют в удобную географическую систему координат $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

в соответствии с выражением (13). Соответствующее значение H_jδ берется из массива цифровой карты рельефа местности.Coordinate Results $$\overrightarrow{V_{f_{j,m,\delta }}}$$

convert to a convenient geographic coordinate system $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

in accordance with the expression (13). The corresponding value of H _jδ is taken from an array of digital terrain maps.

достигается при работе с последовательностью видеокадров и привязанными к ним данными телеметрии (см. фиг.5 и 6). При этом в случае, когда j-й объект неподвижен, на изображении имеется достаточное число контрастных объектов, а нелинейное изменение рельефа в пределах кадра мало по сравнению с высотой ЛПС позволяет упростить действия оператора. В этой ситуации оператор отмечает j-й объект только на первом кадре последовательности, а на последующих кадрах положение объекта определяется автоматически.Improving the accuracy of determining the coordinates of objects $$\overrightarrow{V}=\left(B_j,L_j\right)$$

achieved when working with a sequence of video frames and telemetry data attached to them (see FIGS. 5 and 6). Moreover, in the case when the j-th object is stationary, the image has a sufficient number of contrasting objects, and the nonlinear change in the relief within the frame is small compared to the LPS height allows simplifying the operator’s actions. In this situation, the operator marks the j-th object only on the first frame of the sequence, and on subsequent frames the position of the object is determined automatically.

, можно использовать как для уточнения направления на объект в системе координат видеокамеры для начального положения БЛА $$\overrightarrow{V_{{П}_{ij}}}$$

, так и для непосредственного определения координат объекта интереса $$\overrightarrow{V_j}={\left(B_j,L_j\right)}_l$$

.Coordinates $${\left(x_r,y_r\right)}_l^T$$

, can be used as to clarify the direction to the object in the coordinate system of the camera for the initial position of the UAV $$\overrightarrow{V_{P_{ij}}}$$

, and to directly determine the coordinates of the object of interest $$\overrightarrow{V_j}={\left(B_j,L_j\right)}_l$$

.

In the first embodiment, the random component of the errors in measuring the direction to the object in the coordinate system of the camera is eliminated, and the coordinates of the object are determined only using the coordinates and orientation of the UAVs corresponding to the first frame of the series.

. При этом обеспечивается более высокая точность, поскольку в этом режиме используют результаты измерений координат и ориентации БЛА, сделанные на всей серии кадров. В связи с этим в предлагаемых способе и устройстве реализуется данный подход к определению координат $$\overrightarrow{V_j}$$

.In the second case (being resource-intensive), for each frame of the series, the coordinates of the object of interest are calculated $$\overrightarrow{V_j}$$

. This provides higher accuracy, since in this mode, the results of measurements of the coordinates and orientation of the UAV made on the entire series of frames are used. In this regard, the proposed method and device implements this approach to the determination of coordinates $$\overrightarrow{V_j}$$

.

. Тогда координаты j-го объекта в l+1-м кадре примут вид $${\left(x_r,y_r\right)}_{l+\mathrm{1,}j}^T$$

. Известно (см. Szeliski, Richard. Computer: Algorithms and Applications. Sprintger, 2010), что при съемке плоской поверхности (в данном случае земной поверхности) одной и той же видеокамерой с двух позиций справедливо выражение, связывающее координаты одного и того же объекта на l-м и +1-м кадрах:Write the coordinates of the j-th object in pixels on the l-th frame $${\left(x_r,y_r\right)}_{lj}^T$$

. Then the coordinates of the j-th object in the l + 1st frame will take the form $${\left(x_r,y_r\right)}_{l+\mathrm{one,}j}^T$$

. It is known (see Szeliski, Richard. Computer: Algorithms and Applications. Sprintger, 2010) that when shooting a flat surface (in this case, the earth’s surface) with the same video camera from two positions, the expression connecting the coordinates of the same object on lth and + 1st frames:

$$\left(\begin{array}{c}k{x}_{l+\mathrm{one}}\\ k{y}_{l+\mathrm{one}}\\ k\end{array}\right)=M_{l+\mathrm{one}}\left(\begin{array}{c}{x}_l\\ y_l\\ \mathrm{one}\end{array}\right),\left(\mathrm{fifteen}\right)$$

where М _{l + 1} is the matrix of the projective transformation from the l-th frame to l + 1-th. It is determined by the mutual arrangement of cameras in the l-th and l + 1-th positions and has the form:

. $$M_{l+\mathrm{one}}=\left(\begin{array}{ccc}{m}_{\mathrm{eleven}}& m_{12}& m_{13}\\ m_{21}& m_{22}& m_{23}\\ m_{31}& m_{32}& \mathrm{one}\end{array}\right)$$

.

и $${\left\{\overrightarrow{x}\right\}}_{l+1}$$

. Для каждой точки из набора контрастных точек реализуют вычисление ее характеристики при помощи алгоритма BRIEF (см. М.Colonder, V.Lepetit, С.Strecha, P.Fua. BRIEF: Bonary Robust Independent Elementary Features. ECCV, 2010). Далее путем попарного сравнения вычисленных характеристик точек находят множество пар точек $$\left\{\left(\overrightarrow{x_l},\overrightarrow{x_{l+1}}\right)\right\}$$

, где $$\overrightarrow{x_l}$$

- контрастная точка в l-м изображении (кадре), а $$\overrightarrow{x_l{}_{+1}}$$

- найденная как соответствующая ей контрастная точка в l+1-м изображении.The determination of the value of the matrix M _{l + 1} is as follows. Using the SURF algorithm (see Herbert Bay, Andreas Ess, Tinne Tnytelaars, Luc Van Gool SURF: Speeded Up Robust Features. - Computer Vision and Image Understanding (CVIU), Vol. 110, No. 3, 2008, p. 346-359 ) search for sets of contrast points in two images $${\left\{\overrightarrow{x}\right\}}_l$$

and $${\left\{\overrightarrow{x}\right\}}_{l+\mathrm{one}}$$

. For each point in the set of contrasting points, its characteristics are calculated using the BRIEF algorithm (see M. Colonder, V. Lepetit, C. Strecha, P. Fua. BRIEF: Bonary Robust Independent Elementary Features. ECCV, 2010). Next, by pairwise comparing the calculated characteristics of the points, many pairs of points are found $$\left\{\left(\overrightarrow{x_l},\overrightarrow{x_{l+\mathrm{one}}}\right)\right\}$$

where $$\overrightarrow{x_l}$$

- contrast point in the l-th image (frame), and $$\overrightarrow{x_l{}_{+\mathrm{one}}}$$

- found as the corresponding contrast point in l + 1-m image.

, p=1, 2, …, Р, то согласно (5) имеют место 2Р линейных уравнений относительно восьми неизвестных коэффициентов матрицы М:Upon detection of P pairs of points corresponding to each other $${\left(\overrightarrow{x_l},\overrightarrow{x_{l+\mathrm{one}}}\right)}^p$$

, p = 1, 2, ..., P, then according to (5) there are 2P linear equations for eight unknown coefficients of the matrix M:

$$\begin{array}{l}{x}_{l+\mathrm{one}}^p\left(m_{31}{x}_l^p+m_{32}{y}_l^p+\mathrm{one}\right)=m_{\mathrm{eleven}}{x}_l^p+m_{12}{y}_l^p+m_{13}\\ y_{l+\mathrm{one}}^p\left(m_{31}{x}_l^p+m_{32}{y}_l^p+\mathrm{one}\right)=m_{21}{x}_l^p+m_{22}{y}_l^p+{\mathrm{<figure-callout\; id="23"\; label="m"\; filenames="00000244.png"\; state="\{\{state\}\}">m</figure-callout>}}_{23}.\left(16\right)\end{array}$$

In practice, the number P is large enough compared to the minimum necessary to solve this system. However, among the found correspondences of contrasting points, there may be erroneous coincidences. This problem is resolved using the RANSAC algorithm (see Martin A. Fischler and Robert C. Bolles {June 1981). Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography. Comm. of the ACM14 (6): 381-395. dot 10.1145 / 358669.358692). The algorithm runs at a fixed number of iterations, at each of which s random pairs are selected from n found matches. From the selected pairs, a system of 2s linear equations is obtained, which is solved by the least squares method. For the obtained solution, the discrepancy on the entire set of 2n equations is considered. After performing a fixed number of iterations, the best solution is chosen as M.

на последующем l+1-м кадре последовательности. С каждым последующим кадром местоположение j-го объекта уточняется, что и приводит к повышению точности измерений. Переход от координат $${\left(x_r,y_r\right)}_l^T$$

к вектору направления $$\overrightarrow{V_{ij}}$$

в координатах видеокамеры на каждом очередном кадре осуществляют в соответствии с выражениями (1)-(4).Thus, having found the matrix of projective transformation between adjacent frames, it is possible to find the coordinates of the object $$\overrightarrow{V_j}=\left(B_j,L_j\right){}_{l+\mathrm{one}}$$

on the subsequent l + 1st frame of the sequence. With each subsequent frame, the location of the j-th object is refined, which leads to an increase in the accuracy of measurements. Transition from coordinates $${\left(x_r,y_r\right)}_l^T$$

to direction vector $$\overrightarrow{V_{ij}}$$

in the coordinates of the camera on each next frame is carried out in accordance with the expressions (1) - (4).

в рамках координат видеокамеры, уменьшению случайных ошибок оценивания за счет многократного определения координат объектов по серии кадров, а также благодаря учету особенностей рельефа местности в районе измерений. Достигаемый при этом положительный эффект примерно может быть оценен через отношение площадей засветки земной поверхности видеокамерой к площади объекта (площади кадра к площади пятна засветки от j-го объекта).Thus, in the proposed method for determining the coordinates of objects, an increase in accuracy characteristics is achieved due to a more accurate measurement of the direction vector to the object $$\overrightarrow{V_{P_{ij}}}$$

within the framework of the camera’s coordinates, the reduction of random estimation errors by repeatedly determining the coordinates of objects from a series of frames, and also by taking into account the features of the terrain in the measurement area. The positive effect achieved in this case can be approximately estimated through the ratio of the areas of illumination of the earth's surface by the video camera to the area of the object (the area of the frame to the area of the spot of illumination from the j-th object).

A device for determining the coordinates of objects containing an unmanned aerial vehicle 1 (UAV) and ground control point 2, wherein the UAV 1 is made up of a controller 8, a steering gear 9 and aerodynamic control wheels 11, an autopilot 4, a group of information inputs of which is connected to a second group of information outputs controller 8, the first group of information inputs of which are connected to the group of information outputs of the autopilot 4, a propulsion system 3, the group of information inputs of which is connected to a third the first group of information outputs of controller 8, the first transceiver module 10, the group of information inputs of which is connected to the fourth group of information outputs of the controller 8, the second group of information inputs of which is connected to the group of information outputs of the first transceiver module 10, and the video surveillance unit 5, and ground control point 2 is made comprising serially connected first control unit 13, a second transceiver module 14 and a first information processing and display device 17.

To ensure high-precision measurement of the coordinates of specified objects from the UAV 1, a transmitting module 12, a UAV navigation unit 7 and a storage device 6 are additionally introduced into it; moreover, the first group of information inputs of the storage device 6 is connected to the group of information outputs of the video surveillance unit 5, the second group of information inputs connected to the group of information outputs of the UAV navigation unit 7, and the group of information outputs of the storage device 6 is connected to the group of information inputs of the transmitting mode For 12. At ground control point 2, a receiver module 15 and a second processing and indicating device 16 and a second control unit 18 are additionally introduced in series, the group of information inputs of which is combined with the first group of information inputs of the second information processing and display device 16, and the group of information outputs - with a second group of information inputs of the second device for processing and displaying information 16.

The inventive device for determining the coordinates of objects works as follows (see figure 3). At the preparatory stage, a UAV unit 5 (video camera) is installed on the UAV 1 under the fuselage, for example, an EVS IP camera (see EVS megapixel network cameras.. Evs.ru/prod.php?gr=313). Determine the flight path of the UAV 1 (search for specified objects).

The take-off, flight and landing of the UAV 1 is controlled from the first automated workstation (AWP) of the ground control point 2, consisting of a control unit 13, a second transceiver module 14 and a first information processing and display device 14. This operation is carried out on the first radio channel at a frequency of 0 , 9-0.92 MHz using modules 10 and 14. The UAV control commands from the output of block 13 through the transceiver modules 14 and 10 are received at the input of the controller 8. From the output of block 8, they follow the group of inputs of the motor set lamb 3 and through the steering gear 9 to the aerodynamic steering wheels 11.

Commands supplied to the propulsion system 3 may include control signals for turning the engine 3 on / off, changing the rotational speed of the screw, etc.

The commands given through the steering gear 9 to the aerodynamic steering wheels 11 can change the angles of inclination of the wings, the configuration of their surface and other parameters for controlling the movement of the UAV 1.

Autopilot 4 provides the necessary stabilization of the position of the UAV 1 in space at a height set by the unit 13, parry of wind disturbances, movement along a given route, etc. The impact of the autopilot 4 on the propulsion system 3 and through the steering gear 9 - on the aerodynamic wheels 11 is carried out through the controller 8. The latter generates the necessary control commands for the functional units of the UAV 1 according to the initial data of block 4. It should be noted that the first automated workstation is currently in state simultaneously control the flight of up to four UAVs of the Orlan 10 type.

A video surveillance unit 5, a storage device 6, a UAV navigation unit 7 and a transmitting module 12 take a direct part in measuring the coordinates of objects on board the UAV 1, and in the ground control point 2, a second automated workstation as part of the receiving module 15, the second processing and display device information 16 and the second control unit 18.

If there is data on the position and orientation of the UAV 1, it is possible to determine the coordinates of objects in the video image in an online mode (see Fig. 9). The image obtained by CCTV unit 5 (IP camera EVS), in digital format is transmitted to the NPU 2 using blocks 12 and 15 in the 2.4 GHz band. The data transfer rate is 4 Mbps. When sending data in JPEG format, it is possible to transfer two to four high-resolution frames per second or ten to twelve low-resolution frames per second with a supported resolution of 1600 × 1200 or 640 × 480 (for the EVS camera), respectively. The coordinates of the object are determined at the second automated workstation (blocks 15, 16 and 18), which receives the video stream from block 5 in real time (see Fig. 10). The image obtained on the AWP screen allows the operator to select the specified object using block 18. Thanks to the telemetry data of the UAV 1 and the position of the object in the current frame, it makes it possible with some error to calculate the geographical coordinates of the object of interest.

The functions of block 6 include the joint recording of video frames from block 5 and the corresponding navigation data of UAV1 from the output of block 7.

The performed experimental studies based on the Orlan-10 UAV showed that a number of factors influence the accuracy of determining the coordinates of objects:

the error in determining the geographic coordinates of the UAV with the help of the СРРЯ-receiver is 15-30 m, which entails an error of 15-30 meters;

the error in determining the orientation angles of the Orlan UAV using the built-in accelerometers is currently 1 ° (for small, up to 10 °, UAV deviations 1 from the horizon). The contribution made by this error is proportional to the height H, UAV 1 and for 1000 meters is about 15 m in each of the angles;

the error caused by the delay in the process of linking telemetry to the frame due to delays in the communication channel. When flying at a speed of 100 km / h, a UAV flies 27 meters in one second. The delay of telemetry for one second may introduce an error in the measurements of the order of 25-30 m;

the error caused by the delay in linking telemetry to the frame, caused by the low frequency of updating GPS coordinates (once per second), which also entails an error of 25-30 m.

Thus, the total error in determining coordinates from a height of 1000 m can be 100 m.

In the proposed method and device, measures have been taken to improve the accuracy of measuring coordinates:

more accurate navigation equipment is used, which allows to determine the location of the UAV with an accuracy of 3-5 m, and the orientation of the UAV with an accuracy of tenths of a degree, made in accordance with US Pat. RF №2371733 and №2374659. The measurement error introduced in this case is 2-10 meters;

the use of a digital communication channel in conjunction with the binding of telemetry to video on board the LPS eliminates the error associated with the delay in communication channels;

the use of navigation equipment, which allows updating information on the UAV position up to 10 times per second, can significantly reduce the error associated with linking telemetry to the frame;

averaging the coordinates (B _j , L _j ) or (x _r , y _r ), achieved by their multi-position measurement, makes it possible to reduce the contribution made by measurement errors, which is random in nature.

The combined use of these measures allowed to reduce the total error in determining the coordinates from a height of 1000 meters to 2-10 meters. It should be noted that the largest contribution is made by the error in determining the location of UAVs, which is systematic in nature.

All functional elements and blocks of the proposed device are widely covered in the literature and are commercially available.

As a UAV 1, it is advisable to use the city of St. Petersburg UAV Orlan 10 (see 10.html), which is commercially available from Special Technological Center LLC.

The UAV payload weight is 5 kg, the launch method is from a collapsible catapult, the landing is by parachute. UAV airspeed 90-150 km / h, maximum flight duration - 16 hours, maximum range - 600 km, maximum altitude - 5 km.

CCTV unit 5 can be implemented using a digital IP video camera EVS. The navigation unit 7 may be implemented in accordance with US Pat. RF №2371733 or US Pat. RF №2374659.

UAV 1 "Orlan-10" control is implemented from the first workstation via a low-speed duplex communication channel at frequencies of 900-920 MHz in the pseudo-random tuning of the operating frequency. On this channel (blocks 10 and 14), the flight route, flight altitude and flight order are set: passage at altitude or barrage, etc. Control information is generated using block 13, which can be used as a laptop.

The video image of the objects on NPU 2 from the UAV 1 is supplied via a high-speed simplex channel at frequencies of 2000-2500 MHz to the second AWP. Information transfer speed 4 Mbps. The communication range depends on the flight altitude and local conditions and averages 100-130 km.

The second device for processing and displaying information 16 (see Fig. 11) is designed to determine the coordinates of objects (the decision is made by the operator, the command for the execution of which is generated using block 18), the implementation of operations in accordance with expressions 1-16, the presentation of measurement results in a given form. It contains the first calculator 19, the second calculator 20, the fourth calculator 21, the fifth calculator 23, the sixth calculator 24, the seventh calculator 25, the second storage device 26, the clock generator 27, the eighth calculator 28, the ninth calculator 29, the switching unit 30 , an image processing unit 31, a third control unit 32, a third storage device 33, an averaging unit 34, a comparison unit 35, and an indication unit 36.

The second device for processing and displaying information 16 operates as follows.

At the preparatory stage, using the control unit 18 (as the last, a laptop operating in accordance with the algorithm presented in Fig. 26 can be used), the initial data are set:

orientation of the camera relative to the side (k _k , l _k , ζ _k );

the measured distortion coefficients of the camera lens k ₁ , k ₂ , k ₃ ;

threshold values Δβ, Δd and Δδ;

к вектору направления на него $$\overrightarrow{V_{{П}_{ij}}}$$

;the number of iterations when solving the equation of transition from the coordinates of the object in the frame $${\left(x_r,y_r\right)}^T$$

to the direction vector to it $$\overrightarrow{V_{P_{ij}}}$$

;

the number of calculation iterations in the RANSAC algorithm;

digital map of the measurement area with boundary characteristics of the terrain.

. В функции блока 25 (см. фиг.12 и 15) входит преобразование координат объекта в пикселях $${\left(x_r,y_r\right)}_j^T$$

в вектор направления на него в системе координат видеокамеры 5 $$\overrightarrow{V_{{П}_{ij}}}$$

в соответствии с выражениями 1-4.During operation, the video image is read by blocks 18 and 31. When a given object is detected, information about it from the output of block 18 is sent to the first group of information inputs of the seventh calculator 25 in the form of coordinates $${\left(x_r,y_r\right)}_j^T$$

. The functions of block 25 (see Fig. 12 and 15) include the transformation of the coordinates of the object in pixels $${\left(x_r,y_r\right)}_j^T$$

into the direction vector to it in the coordinate system of the camera 5 $$\overrightarrow{V_{P_{ij}}}$$

in accordance with expressions 1-4.

и $${\beta }_{lp{s}_i}$$

поступает на группу информационных входов первого вычислителя 19. В его функции входит преобразование пространственных параметров БЛА $$\text{V}\overrightarrow{{{}_{\text{lps}}}_i}$$

в геоцентрическую систему координат $$\overrightarrow{V_{lp{s}_i}^{Г}}={(X_{lps},Y_{lps},Z_{lps})}_i$$

) в соответствии с выражением (5).At the same time, information about the spatial position of the UAV (B _lps , L _lps , H _lps ) _i , as well as $${\theta }_{lp{s}_i}$$

and $${\beta }_{lp{s}_i}$$

arrives at the group of information inputs of the first calculator 19. Its function includes the transformation of the spatial parameters of the UAV $$\text{V}\overrightarrow{{{}_{\text{lps}}}_i}$$

to geocentric coordinate system $$\overrightarrow{V_{lp{s}_i}^G}={(X_{lps},Y_{lps},Z_{lps})}_i$$

) in accordance with expression (5).

поступают на первую группу информационных входов второго вычислителя 20, а на вторую группу его информационных входов - значение $$\overrightarrow{V_{{П}_{ij}}}$$

с группы информационных выходов блока 25. В функции вычислителя 20 входит коррекция вектора направления на j-й объект $$\overrightarrow{V_{{П}_{ij}}}$$

на основе априорно известной ориентации видеокамеры 5 относительно борта БЛА 1. Последняя поступает с первой группы информационных входов устройства 16 на входы второго запоминающего устройства 26, представляющего собой буферное запоминающее устройство. С информационных выходов блока 26 значения (k_k, l_k, ζ_k) следуют на третью группу информационных входов второго вычислителя 20. Скорректированный вектор $$\overrightarrow{V_{{}_{ij}}^{\text{'}}}$$

находят путем последовательного умножения вектора $$\overrightarrow{V_{{П}_{ij}}}$$

на три соответствующие углам Эйлера матрицы поворота в соответствии с (6).Calculation results $$\overrightarrow{V_{lp{s}_i}^G}$$

enter the first group of information inputs of the second computer 20, and the second group of its information inputs - value $$\overrightarrow{V_{P_{ij}}}$$

from the group of information outputs of block 25. The functions of the calculator 20 include the correction of the direction vector to the jth object $$\overrightarrow{V_{P_{ij}}}$$

based on the a priori known orientation of the video camera 5 relative to the side of the UAV 1. The latter comes from the first group of information inputs of the device 16 to the inputs of the second storage device 26, which is a buffer storage device. From the information outputs of block 26, the values (k _k , l _k , ζ _k ) follow the third group of information inputs of the second calculator 20. The adjusted vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

found by sequentially multiplying the vector $$\overrightarrow{V_{P_{ij}}}$$

into three rotation matrices corresponding to Euler angles in accordance with (6).

с выходов блока 20 поступают на первую группу информационных входов третьего вычислителя 21. В его функции входит определение уточненного вектора направления на j-й объект $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}={\left(X_{ij}^{+},Y_j^{+},Z_j^{+}\right)}_i$$

с учетом измеренных в момент времени t_i пространственных углов БЛА: крена $${\text{k}}_{\text{lps}}{}_i$$

, тангажа $${\text{l}}_{\text{lps}}{}_i$$

, курсового угла $${\alpha }_{\text{lps}}{}_i$$

и склонения $${\zeta }_{\text{lps}}{}_i$$

. Для этого используют нормальную систему координат. С учетом вышесказанного на вторую группу информационных входов блока 21 подают значения ( $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

), характеризующие ориентацию БЛА в пространстве в момент получения кадра видеоизображения с j-м объектом. Определение уточненного вектора направления $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

в блоке 21 выполняют в соответствии с выражением (7).Adjusted Vector Values $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

from the outputs of block 20 go to the first group of information inputs of the third calculator 21. Its function is to determine the updated direction vector to the j-th object $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}={\left(X_{ij}^{+},Y_j^{+},Z_j^{+}\right)}_i$$

taking into account the measured at time t _{i the} spatial angles of the UAV: roll $${\text{k}}_{\text{lps}}{}_i$$

pitch $${\text{l}}_{\text{lps}}{}_i$$

heading angle $${\alpha }_{\text{lps}}{}_i$$

and declensions $${\zeta }_{\text{lps}}{}_i$$

. To do this, use the normal coordinate system. In view of the above, the values of ( $${\text{k}}_{\text{lps}}{}_i$$

, $${\text{l}}_{\text{lps}}{}_i$$

, $${\zeta }_{\text{lps}}{}_i$$

), characterizing the orientation of the UAV in space at the time of receipt of the video frame with the j-th object. Definition of a refined direction vector $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

in block 21 is performed in accordance with the expression (7).

с информационных выходов блока 21 далее следует на первую группу информационных входов четвертого вычислителя 22. В функции последнего входит определение уточненных значений азимутального угла $${\theta }_{ij}^{+}$$

, угла места $${\beta }_{ij}^{+}$$

и удаления j-го объекта от БЛА d(H₀)_ij. Пространственные углы $${\theta }_{ij}^{+}$$

и $${\beta }_{ij}^{+}$$

находят в соответствии с выражениями (8) и (9) соответственно. Расстояние d(H₀)_ij между БЛА и j-м объектом определяют в соответствии с (10). Для обеспечения вычислений на вторую группу информационных входов блока 22 поступает значение $$H_{\text{lps}}{}_i$$

с первой группы информационных входов устройства 16. Радиус Земли известен, а его значение содержится в блоке 22. В случае невозможности определить расстояния d(H₀)_ij на выходе обнуления четвертого вычислителя 22 формируется сигнал, который поступает на входы обнуления первого 19, второго 20 и третьего 21 вычислителей. В результате значения векторов $$\overrightarrow{V_{{П}_{ij}}}$$

, $$\overrightarrow{V_{{}_{ij}}^{\text{'}}}$$

и $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

в этих блоках обнуляются, а заявляемое устройство начинает новый цикл работы.The adjusted value of the direction vector to the jth object $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

from the information outputs of block 21, it then goes to the first group of information inputs of the fourth calculator 22. The function of the latter includes determining the updated values of the azimuthal angle $${\theta }_{ij}^{+}$$

elevation $${\beta }_{ij}^{+}$$

and removing the j-th object from the UAV d (H ₀ ) _ij . Spatial angles $${\theta }_{ij}^{+}$$

and $${\beta }_{ij}^{+}$$

find in accordance with expressions (8) and (9), respectively. The distance d (H ₀ ) _ij between the UAV and the j-th object is determined in accordance with (10). To provide calculations, the second group of information inputs of block 22 receives a value $$H_{\text{lps}}{}_i$$

from the first group of information inputs of device 16. The radius of the Earth is known, and its value is contained in block 22. If it is impossible to determine the distances d (H ₀ ) _ij at the zeroing output of the fourth calculator 22, a signal is generated that goes to the zeroing inputs of the first 19, second 20 and the third 21 calculators. As a result, the values of the vectors $$\overrightarrow{V_{P_{ij}}}$$

, $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

and $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

in these blocks are reset, and the inventive device begins a new cycle of work.

с группы информационных выходов блока 21 через блок 22 поступает на группу информационных входов пятого вычислителя 23. В его функции входит преобразование уточненного вектора направления на j-й объект, находящегося в нормальной системе координат, в истинный вектор направления $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

в геоцентрической системе координат. Данную операцию в блоке 23 осуществляют в соответствии с выражением (11). Для этого на вторую группу информационных входов пятого вычислителя 23 подают значения $${\text{B}}_{\text{lps}}{}_i$$

и $${\text{L}}_{\text{lps}}{}_i$$

с первой группы информационных входов блока 16.To measure d (H ₀ ) _{ij the} value $$\overrightarrow{V_{{}_{{}_{ij}}}^{+}}$$

from the group of information outputs of block 21 through block 22, it goes to the group of information inputs of the fifth calculator 23. Its function is to convert the refined direction vector to the jth object in the normal coordinate system into a true direction vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

in a geocentric coordinate system. This operation in block 23 is carried out in accordance with the expression (11). To do this, the second group of information inputs of the fifth transmitter 23 serves the values $${\text{B}}_{\text{lps}}{}_i$$

and $${\text{L}}_{\text{lps}}{}_i$$

from the first group of information inputs of block 16.

. Последний поступает на вторую группу информационных входов шестого 22 и пятую группу информационных входов восьмого 28 вычислителей и вторую группу информационных входов третьего блока управления 32.At the next stage of operation of the inventive device, the vector is converted $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

. The latter goes to the second group of information inputs of the sixth 22 and the fifth group of information inputs of the eighth 28 calculators and the second group of information inputs of the third control unit 32.

с "круглой" Землей $$\overrightarrow{V_{фj}}$$

и преобразовании геоцентрических координат j-го объекта в географические $$\overrightarrow{V_j}$$

. Первую из названных функций в блоке 24 выполняют в соответствии с выражением (12). Для этого значение $$\overrightarrow{V_{lp{s}_i}^{Г}}$$

, сформированное первым вычислителем 19, поступает на первую группу информационных входов шестого вычислителя 24. Кроме того, значение d(H₀)_ij, найденное блоком 22, через блок 23 поступает на вторую группу информационных входов 24. Далее осуществляют переход от геоцентрического вектора координат j-го объекта к его географическим координатам В_j и L_j в соответствии с (13). Результаты вычислений $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

с выходов блока 24 следуют на вторую группу информационных входов блока коммутации 30.The purpose of the sixth calculator 24 is to determine the coordinates of the point of intersection of the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

with round earth $$\overrightarrow{V_{fj}}$$

and transforming the geocentric coordinates of the j-th object into geographical $$\overrightarrow{V_j}$$

. The first of these functions in block 24 is performed in accordance with expression (12). For this value $$\overrightarrow{V_{lp{s}_i}^G}$$

generated by the first calculator 19, goes to the first group of information inputs of the sixth calculator 24. In addition, the value d (H ₀ ) _ij found by block 22, through block 23 goes to the second group of information inputs 24. Next, the transition from the geocentric coordinate vector j -th object to its geographical coordinates B _j and L _j in accordance with (13). Calculation results $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

from the outputs of block 24 are followed by a second group of information inputs of the switching unit 30.

на предварительном Δd и конечном Δδ этапах. В блок 33 осуществляют упорядоченную (по заданным адресам) запись цифровой карты рельефа местности. Матрица охватывает участок земной поверхности, ограниченный координатами (В_а, L_а) и (B_b,L_b). Назначение блока управления 32 состоит в преобразовании части вектора $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

, ограниченного точками ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) и (B_j, L_j) в линейку адресов {A_i,j,δ}, соответствующих равномерно распределенным по его длине высотам {H_i,j,δ} рельефа местности. С этой целью на первую группу информационных входов блока 32 поступают координаты БЛА 1 ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

). На вторую группу информационных входов блока управления 32 подают значение вектора направления на j-й объект $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

с выходов блока 23. На третьей группе информационных входов блока управления 32 присутствует значение координат $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

, поступившее с выхода шестого вычислителя 24. В блоке 32 названный отрезок вектора $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

преобразуют в последовательность адресов {A_i,j,δ}, которые поступают на адресные входы запоминающего устройства 33. Последние используются для формирования на его входе адресной линейки A_i,j,δ. В результате на третью группу информационных входов восьмого вычислителя 28 поступает последовательность высот {H_i,j,δ} рельефа местности, соответствующая заданному отрезку истинного вектора направления $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}}}$$

. Емкость последовательности высот определяется значением М=max{J,δ}, где J=d(H₀)_ij/Δd, δ=Δd/Δδ. На фиг.17 приведен алгоритм работы блока 28 по поиску предварительного и точного (с заданной точностью) определения координат объектов $$\overrightarrow{V_{фj\delta }}$$

.At the same time (with block 22) in block 28 in two stages, the coordinates of the j-th object are determined with a given accuracy. This operation is performed in conjunction with the control unit 32 and the storage device 33. At the preparatory stage, a digital terrain map of the measurement area is recorded in the storage device 33. This operation is performed using block 18 along the first group of information inputs of block 16. At the same time, the boundary values of the DEM (B _a , L _a ) and (B _b , L _b ) and the number of split points J, and in block 28, the number of scanning steps along the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

at the preliminary Δd and final Δδ stages. In block 33 carry out an ordered (at given addresses) recording a digital map of the terrain. The matrix covers a portion of the earth's surface, limited by the coordinates (B _a , L _a ) and (B _b , L _b ). The purpose of the control unit 32 is to convert part of the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

bounded by points ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) and (B _j , L _j ) to the line of addresses {A _{i, j, δ} } corresponding to the heights {H _{i, j, δ} } of the terrain that are uniformly distributed along its length. For this purpose, the coordinates of the UAV 1 ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) The second group of information inputs of the control unit 32 serves the value of the direction vector to the j-th object $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

from the outputs of block 23. On the third group of information inputs of the control unit 32 there is a coordinate value $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

received from the output of the sixth calculator 24. In block 32, the named segment of the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

transform into a sequence of addresses {A _{i, j, δ} }, which are fed to the address inputs of the storage device 33. The latter are used to form the address line A _{i, j, δ} at its input. As a result, the third group of information inputs of the eighth transmitter 28 receives a sequence of heights {H _{i, j, δ} } of the terrain corresponding to a given segment of the true direction vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

. The capacity of the sequence of heights is determined by the value M = max {J, δ}, where J = d (H ₀ ) _ij / Δd, δ = Δd / Δδ. On Fig shows the algorithm of the block 28 to search for preliminary and accurate (with a given accuracy) determination of the coordinates of objects $$\overrightarrow{V_{fj\delta }}$$

.

в соответствии с выражением (14).The geocentric coordinates of the j-th object are then fed to the information inputs of the ninth calculator 29. In block 29, the geocentric coordinates are converted into geographical $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

in accordance with expression (14).

The results of the calculations from the output of block 29 go to the first group of information inputs of the switch 30.

, полученные на "круглой" Земле или точные $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

с учетом рельефа местности) принимает блок сравнения 35. На подготовительном этапе (с использованием блока 18) в блок сравнения 35 записывают значение Δβ, определяющее заданную потенциальную точность определения координат объектов. В процессе работы заявляемого устройства в блоке 35 выполняют сравнение очередного измеренного значения $${\beta }_{ij}^{+}$$

с пороговым значением Δβ. Если текущее значение $${\beta }_{ij}^{+}$$

оказалось меньше порогового уровня Δβ, блок 35 формирует управляющий сигнал, поступающий на вход управления блока коммутации 30. В результате значение координат $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

с выхода блока 24 через блок 30 поступает на группу информационных входов блока усреднения координат 34. В противном случае ( $${\beta }_{ij}^{+}>\Delta \beta$$

), а на вход блока 34 поступает значение $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

с выхода блока 29.The decision about which coordinates will go to the input of block 34 (approximate $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

obtained on round Earth or accurate $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

taking into account the terrain) receives the comparison unit 35. At the preparatory stage (using block 18), the value Δβ, which determines the specified potential accuracy of determining the coordinates of objects, is written to the comparison unit 35. In the process, the inventive device in block 35 compares the next measured value $${\beta }_{ij}^{+}$$

with a threshold value Δβ. If the current value $${\beta }_{ij}^{+}$$

turned out to be less than the threshold level Δβ, block 35 generates a control signal fed to the control input of switching unit 30. As a result, the coordinate value $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

from the output of block 24 through block 30, it enters the group of information inputs of the coordinate averaging block 34. Otherwise ( $${\beta }_{ij}^{+}>\Delta \beta$$

), and the input of block 34 receives the value $$\overrightarrow{V_{j\delta }}=\left(B_j,L_j,H_j\right)$$

from the output of block 29.

), блок 28 работает по алгоритму (см. фиг.17) в соответствии с выражением (13), а на вход блока 34 поступают приблизительные координаты j-го объекта $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

с выхода блока 24. Синхронность выполнения всех операций обеспечивает генератор синхроимпульсов 27.If a situation arises in which there is no information about the terrain and ( $${\beta }_{ij}^{+}>\Delta \beta$$

), block 28 operates according to the algorithm (see Fig. 17) in accordance with expression (13), and the approximate coordinates of the j-th object arrive at the input of block 34 $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

from the output of block 24. Synchronization of all operations is provided by the clock generator 27.

When evaluating the coordinates of one or the first frame of the video image of the object at the output of block 31, there is no control signal. In block 34, the averaging operation is not performed, and the coordinates of the j-th object are input to the display unit 36. In addition, the measurement results are stored in its buffer memory. This ends the stage of operation of the device for one frame of the video image.

At the next stage (at the next frame), using the video image analysis unit 31, the received frame is analyzed for the presence of an image of a given j-th object in it. However, the operation algorithm of the device changes when it detects a j-th object previously observed in a subsequent frame. In this case, multiple (by the number of frames with the image of the object) measurements are made of its coordinates, followed by their averaging in block 34, which can significantly improve the accuracy of measurements. The main element that implements these measurements is the image processing unit 31. If the j-th object is not detected by block 31 on the next frame, the results of measurements of its averaged coordinates remain unchanged, they are sent to the information inputs of the indicating unit 36.

When j + 1-th object is detected, all the above operations are repeated.

Block 31 is designed to solve the following problems (see Fig. 14 and 16):

search for contrast points in the image using the SURF method;

calculating the characteristics of contrast points using the BRIEF method;

comparison with contrasting points of the previous frame and deciding on the presence of a given object in the current frame;

upon repeated detection of a given object - to generate a control signal to block 34 about the need to average the measurement results;

j-го объекта на l+1-м кадре.the formation of the matrix of projective transformation by the RANSAC method with subsequent refinement of coordinates $${\left(x_r,y_r\right)}_j^T$$

j-th object on l + 1st frame.

или $$\overrightarrow{V_{j\delta }}$$

с последующим усреднением результатов. При первом пропадании изображения объекта в кадре результаты усреднения его координат фиксируются (предыдущие измерения стираются), а в блоке индикации 36 высвечиваются усредненные координаты j-го объекта.Block 31 performs these functions in accordance with expressions (15) and (16). If a decision is made by block 31 that there is an image of the j-th object that was previously present on the 1st frame in the next l + 1th frame, the latter generates a control signal to block 34. As a result, the coordinates of the object obtained at l + 1 -th frame, are stored together with the results obtained on the l-th frame. Next, in the block, the operation of averaging coordinates over two frames is performed, and the results are displayed in display unit 36. The presence of the j-th object in all subsequent frames leads to the next measurement of its coordinates $$\overrightarrow{V_j}$$

or $$\overrightarrow{V_{j\delta }}$$

followed by averaging the results. When the image of the object disappears for the first time in the frame, the results of averaging its coordinates are fixed (previous measurements are erased), and the averaged coordinates of the j-th object are displayed in display unit 36.

в геоцентрической системе координат, а также $${\theta }_{ij}^{+}$$

и $${\beta }_{ij}^{+}$$

и удаление объекта от БЛА d(H₀)_ij (см. фиг.11). Это достигается благодаря учету ориентации видеокамеры относительно борта БЛА1 и определению места объекта в кадре видеокамеры 5 и собственно угловой ориентации БЛА 1 в пространстве. Каждый из вычислителей выполняет строго определенные в выражениях (1-11) операции, реализация которых сложностей не вызывает. Реализация этих блоков известна (см. пат. РФ №2458360, опубл. 10.08.2012 г.), выполняются на постоянных запоминающих устройствах К541 и К500 сериях микросхем. Алгоритмы работы вычислителей 19, 20, 21, 22 и 23 приведены на фиг.18-22 соответственно, а седьмого вычислителя 25 на фиг.15. Кроме того, на фиг.12 и 13 приведен вариант реализации блока 25 на дискретных элементах в соответствии с выражениями (1-4). Для уменьшения массогабаритных характеристик потребляемого тока блоки с 19 по 23 и 25 целесообразно реализовать на специализированном процессоре 7MS320c6416 (см. TMS320c6416: http://fociis/ti/com/docs/prod/folders/print/TMS320c6416.html).The first 19, second 20, third 21, fourth 22, fifth 23 and seventh 25 calculators are designed to determine the true direction vector to the j-th object $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

in the geocentric coordinate system as well $${\theta }_{ij}^{+}$$

and $${\beta }_{ij}^{+}$$

and the removal of the object from the UAV d (H ₀ ) _ij (see Fig.11). This is achieved by taking into account the orientation of the camera relative to the side of the UAV1 and determining the location of the object in the frame of the video camera 5 and the actual angular orientation of the UAV 1 in space. Each of the calculators performs operations strictly defined in expressions (1-11), the implementation of which does not cause difficulties. The implementation of these blocks is known (see US Pat. RF No. 2458360, publ. 08/10/2012), are performed on permanent memory devices K541 and K500 series of microcircuits. The operation algorithms of the calculators 19, 20, 21, 22, and 23 are shown in Figs. 18-22, respectively, and the seventh calculator 25 in Fig. 15. In addition, Figs. 12 and 13 show an embodiment of a block 25 on discrete elements in accordance with expressions (1-4). To reduce the weight and size characteristics of the current consumption, it is advisable to implement blocks 19 through 23 and 25 on a specialized processor 7MS320c6416 (see TMS320c6416: http: //fociis/ti/com/docs/prod/folders/print/TMS320c6416.html).

с "круглой" Землей $$\overrightarrow{V_{фj}}$$

и преобразования геоцентрических координат $$\overrightarrow{V_{фj}}$$

в географические $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

в соответствии с (12) и (13).Sixth 24, eighth 26 and ninth 29 calculators are implemented similarly to the corresponding blocks in US Pat. RF №2458360, publ. 08/10/2012, the Sixth calculator 24 is designed to determine the coordinates of the point of intersection of the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

with round earth $$\overrightarrow{V_{fj}}$$

and transformations of geocentric coordinates $$\overrightarrow{V_{fj}}$$

into geographical $$\overrightarrow{V_j}=\left(B_j,L_j\right)$$

in accordance with (12) and (13).

The implementation of the block of difficulties does not cause. It can be implemented on permanent memory devices K541 and K500 series of microcircuits. The operation algorithm is shown in Fig.23.

. Данную функцию блок 28 выполняет в два этапа в соответствии с алгоритмом, приведенным на фиг.17, и выражением (14). Может быть реализован по аналогии с соответствующим блоком (см. пат. РФ №2458360, опубл. 10.08.2012 г.) на базе 16-ти разрядного микропроцессора К1810ВМ86.The eighth transmitter 28 is designed to determine the location of an object with a given accuracy in a geocentric coordinate system $$\overrightarrow{V_{fj\delta }}$$

. Block 28 performs this function in two stages in accordance with the algorithm shown in Fig. 17 and expression (14). It can be implemented by analogy with the corresponding unit (see US Pat. RF No. 2458360, published on 08/10/2012) based on the 16-bit K1810VM86 microprocessor.

в географические $$\overrightarrow{V_{j\delta }}$$

в соответствии сThe ninth calculator 29 is designed to convert the geocentric coordinates of the object $$\overrightarrow{V_{fj\delta }}$$

into geographical $$\overrightarrow{V_{j\delta }}$$

in accordance with

algorithm in Fig.24 and the expression (13). The implementation of block 29 is known and does not cause difficulties. Block 29 can be implemented on discrete elements based on TTL signal levels, for example 555, 1533 series of microcircuits, etc.

, ограниченной точками ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) и (B_j, L_j), в линейку адресов {A_i,j,δ} соответствующих равномерно распределенным высотам {H_i,j,δ} рельефа местности. Реализация блока 32 известна и трудностей не вызывает. Может быть реализован на микропроцессорной сборке с достаточным быстродействием (см. Шевкоплес Б.В. Микропроцессорные структуры. Инженерные решения: Справочник. - 2-е изд., перераб. и доп. - М.: Радио и связь, 1990. -512 с.), в которой реализован алгоритм, приведенный на фиг.25. Последний определяет порядок выполнения операций по предварительному определению координат объектов. В режиме измерения координат с заданной точностью порядок работы блока 32 сохраняется (алгоритм имеет аналогичный вид).The third control unit 32 is designed to convert part of the vector $$\overrightarrow{V_{{}_{ij}}^{\text{'}\text{'}\text{'}\text{'}}}$$

bounded by points ( $${\text{B}}_{\text{lps}}{}_i$$

, $${\text{L}}_{\text{lps}}{}_i$$

) and (B _j , L _j ), in the line of addresses {A _{i, j, δ} } corresponding to uniformly distributed heights {H _{i, j, δ} } of the terrain. The implementation of block 32 is known and does not cause difficulties. It can be implemented on a microprocessor assembly with sufficient speed (see Shevkoples B.V. Microprocessor structures. Engineering solutions: Reference book. - 2nd ed., Revised and additional - M .: Radio and communications, 1990. -512 s .), which implements the algorithm shown in Fig.25. The latter determines the order of operations for the preliminary determination of the coordinates of objects. In the coordinate measurement mode with a given accuracy, the operation order of block 32 is preserved (the algorithm has a similar form).

The image processing unit 31 (see FIGS. 14 and 16) performs its functions in accordance with expressions (15) and (16). Block 31 contains serially connected block for searching for contrast points 49, block for calculating characteristics of contrast points 50, storage device 53, analysis unit 51, initial data generator 52, statistical processing unit 55, multiplier 56 and divider 57, the group of information outputs of which is the first group of information the outputs of the image processing unit 31, the first group of information inputs of which are connected to the group of information outputs of the block for searching contrast points 49, and the second group of information inputs s is connected to the second group of information inputs of the multiplier 56, the second group of information outputs of the block for calculating the characteristics of contrast points 50 is connected to the second group of information inputs of the analysis unit 51, and a random number sensor 54, the group of information inputs of which is connected to the second group of information inputs of the statistical processing unit, and the clock inputs of all blocks 49 to 57 are combined and connected to the synchronization input of the image processing unit 31.

.The video image from the group of information outputs of the block 15 goes to the group of information inputs of the block for searching for contrast points 49. The function of block 49 includes the implementation of the well-known SURF algorithm. The detected contrast points on the l + 1-th frame with a description of their relative location are fed to the group of inputs of the block for calculating the characteristics of the contrast points 50, which are performed in accordance with the BRIEF algorithm. The latter's task is to calculate bit vectors (sets of bits) that describe the distribution of image brightness in the vicinity of a given contrast point $${\left(x_r,y_r\right)}^T$$

.

The characteristics of the contrasting points with a description of their relative location simultaneously go to the input groups of the storage device 53 (for use in the next frame) and the analysis unit 51. The function of the block 51 includes comparing the images on the previous l-th (received from the outputs of block 53) with the current l + 1 frames. This operation is performed on contrasting points: their relative position on the frames and characteristics. A description of the matching pairs of contrasting points from the l-th and l + 1-th frames is sent to the group of information outputs of the source data former 52. Block 52 provides the transformation of the data received from block 51 to the form necessary for the normal operation of block 55.

Statistical processing unit 55 is intended to solve the system of equations (16). Block 55 performs this function on the basis of the well-known RANSAC algorithm and data arriving at its first group of information inputs from the group of outputs of block 52. In the process, a sequence of random numbers generated by block 54 is used. Solving the above system of equations allows one to determine the coefficients of the desired projective transformation matrix.

, что позволяет получить его координаты на l+1-м кадре (см. выражение (15)). Значение вектора $${\left(x_r,y_r\right)}_l^T$$

поступает со второй группы информационных выходов блока 25. Блок 57 предназначен для деления полученного вектора координат на константу К (его третий элемент z).In block 56, the projective transformation matrix found by block 55 is multiplied by the object coordinate vector on the lth frame $${\left(x_r,y_r\right)}_l^T$$

, which allows us to obtain its coordinates on the l + 1st frame (see expression (15)). Vector value $${\left(x_r,y_r\right)}_l^T$$

comes from the second group of information outputs of block 25. Block 57 is designed to divide the resulting coordinate vector by a constant K (its third element z).

Block 31 can be implemented on a microprocessor assembly with sufficient speed (see Shevkoples B.V. Microprocessor structures. Engineering solutions: Handbook. - 2nd ed., Revised and additional - M .: Radio and communications, 1990. - 512 s.), In which the algorithm of FIG. 16 is implemented.

It is advisable to implement the functions of blocks 24-35 using a second signal processor (see TMS320c6416: http: //focus/ti/com/docs/prod/folders/print/TMS320c6416.html).

In addition, blocks 16 and 18 can be simultaneously implemented on a personal computer. As the minimum requirements for it, you can determine the following: Core i5 processor 2000 MHz, 1 GB of RAM, 200 MB of free space on your hard drive. Software component: operating system Windows XP SP2 and higher, library. NetFrameWork 4.0, a digital terrain map with terrain information and a format compatible with Panorama Group maps.

Claims (5)

1. The method of determining the coordinates of objects, which consists in the fact that at the preparatory stage, a video camera is installed under the fuselage on board the flight lifting equipment (LPS), the orientation of the camera relative to the LPS side (k _k , l _k , ζ _k ) is determined, where k _k , l _k , ζ _k are the angles of roll, pitch and declination of the camera, respectively, and during the flight, the location of the LPS is constantly determined at a given time interval Δt

Where

respectively, the latitude, longitude and height of the LPS, and its spatial orientation

Where

accordingly, the angles of roll, pitch and declination of the LPS at the i-th moment of time, jointly remember the navigation and time parameters of the LPS, and when the j-th given object is visually detected at the time t _i , the direction vector to it is preliminarily determined

in the coordinate system of the video camera, translate the LPS coordinates into a geocentric coordinate system, adjust the direction vector to the j-th object

taking into account the a priori known orientation of the camera
relative to the LPS board (k _k , l _k , ζ _k ) by sequentially multiplying the values

to the rotation matrix corresponding to the Euler angles, after which, in the normal coordinate system, the updated value of the direction vector is calculated

to the j-th object, taking into account the measured at the time t _i spatial angles of the LPS: roll

pitch

, and declensions

, determine the specified azimuth values

elevation

and removal of LPS located at time t _i at a height

, from the jth object

located on the surface of the "round" Earth, in the geocentric coordinate system determine the value of the true direction vector to the j-th object

which depends on latitude B _lps , longitude L _lps , LPS location, determine the coordinates of the intersection point of the vector

with round earth

transform the geocentric coordinates of the j-th object

into geographical

where B _j and L _j, respectively, the latitude and longitude of the location of the j-th object, characterized in that the distortion coefficients of the camera lens k ₁ , k ₂ and k ₃ are pre-measured and stored, the position of the camera relative to the LPS side is fixed for the entire measurement period, as the j-th object can be any stationary or moving physical object observed in the video camera, and the operator decides whether to measure the coordinates of the observed object, the value of the preliminary direction vector to the j-th object

determined by the location of the object on the frame at time t _i, specify the value of the preliminary direction vector to the j-th object

by eliminating the influence on the distortion results of the video camera, if there are n consecutive frames, n = 2, 3, ..., N, n cycles of measurements of geographical coordinates are performed with the image of the j-th object

and the measurement results are averaged, in the presence of a digital map of the terrain of the measurement region, which is a matrix with a given discreteness in the coordinates of the measurement region with the corresponding elevation heights, the geographical coordinates of the detected j-th object are additionally specified

2. The method according to claim 1, characterized in that the coordinates of the j-th object

in the frame is determined in pixels, counted from the upper left corner of the frame of the video camera. 3. The method according to claim 1, characterized in that the transition from the coordinates of the j-th object in the frame to the refined preliminary vector of direction to it

in the coordinate system of the camera, carried out in accordance with the expression

Where

are the coordinates of the center of the matrix (frame) in pixels, ƒ is the focal length of the camera lens, converted into pixels of the matrix,

k ₁ , k ₂ , k ₃ - measured lens distortion coefficients using the simple iteration method. 4. The method according to claim 1, characterized in that in the presence of a digital elevation map of the terrain of the measurement region, a sequential set of elevation values {H _{i, m} }, m = 1, 2, ..., M is formed, which corresponds to uniformly distributed coordinates on a segment, connecting coordinates

and (B _j , L _j }, M = d (H ₀ ) / Δd, where Δd is the scanning step along the direction vector

on the j-th object, determined by the specified accuracy of the preliminary measurement of the coordinates of the object, calculate the coordinates

corresponding to discrete selected elevations of the terrain H _{i, m} , and for the preliminary coordinates of the j-th object

take the first split point of the vector

located below the level of the terrain, specify the location of the j-th object by highlighting the neighboring break point

located above the terrain, segment

direction vector to j-th object

divided into δ equal intervals, Δδ << Δd, where Δδ is the scanning step along the selected segment of the direction vector

and is determined by the final specified accuracy of measuring the coordinates of objects, for the named points, coordinates are calculated

and the corresponding values of the height of the terrain H _{i, m, δ} , for the exact coordinates of the j-th object take the value

located between neighboring points located above and below the terrain, and the obtained value of the coordinates of the j-th object

convert to a convenient geographic coordinate system

5. A device for determining the coordinates of objects, consisting of an unmanned aerial vehicle (UAV) and ground control station, and the UAV is made containing serially connected controller, steering gear and aerodynamic steering wheels, an autopilot, the group of information inputs of which are connected to the second group of information outputs of the controller, the first group the information inputs of which are connected to the group of information outputs of the autopilot, a propulsion system, the group of information inputs of which is connected to the third a group of information outputs of the controller, the first transceiver module, the group of information inputs of which are connected to the fourth group of information outputs of the controller, the second group of information inputs of which is connected to the group of information outputs of the first transceiver module, and a video surveillance unit, and the ground control station is made containing connected to the first control unit for generating UAV take-off, flight and landing control commands, the second transceiver the first module and the first device for processing and displaying information, characterized in that the UAV additionally includes a transmitting module, a UAV navigation unit and a storage device, moreover, the first group of information inputs of the storage device is connected to the group of information outputs of the video surveillance unit, the second group of information inputs is connected to a group of information outputs of the UAV navigation unit, and a group of information outputs of the storage device is connected to a group of information inputs of the transmitting module, in addition, a ground-mounted receiving module and a second processing and indication device, a second control unit designed to set the initial data and form a command to determine the coordinates of objects, the group of information inputs of which are combined with the first group of information inputs of the second device for processing and displaying information, are additionally introduced into the ground control station and the group of information outputs with the second group of information inputs of the second device for processing and displaying information .

Images