RU2142144C1 - Gear determining coordinates and color of object - Google Patents

Abstract

FIELD: radio engineering. SUBSTANCE: gear is designed for usage in systems of visual sensing of robots, in devices determining deformation degree of parts and structures, in systems of automation of study of materials. Gear includes former of present image, masks of basic and additional image, unit of correlation-extreme processing of information, mechanism changing basic and additional standard images, electrooptical filter, unit changing color and unit isolating frame and line synchronization pulses. EFFECT: increased precision of determination of coordinates and color of object. 14 dwg, 1 tbl

Description

 The invention relates to the field of radio engineering and can be used to create devices designed to generate visual information of the robot or to determine the degree of deformation of the surface of parts and structures. The use of the device in materials science makes it possible to automate research, which undoubtedly will not only increase their efficiency, but will also free some of the staff from intense monotonous work. Determining the color of an object provides great opportunities for material scientists, in whose research the color of the structural elements of the surface of the material is one of the main informative features.

 A device for determining the location (coordinates) of an object, comprising a television image sensor, a current image indicator, an optical correlator, a television transmitting tube, a storage device, a comparison circuit, a program unit, a tape drive, a reference image, a horizontal scanning unit, a control unit, coordinate calculator, frame number counter, actuator, filter unit, scaling unit, reverser, displacement calculator, speed calculator growth, digital indication unit, video amplifier, electronic key [1, Fig. 5.7, p. 190].

 The disadvantage of this device is the lack of accuracy in determining the coordinates of the object (for example, a manipulation object (OM) located in the working area of the robot, or structural elements (SE) of the material surface).

 The closest technical solution to the invention is a device containing the imager of the current image, an additional reference image, a block of correlation-extreme information processing, a pulling mechanism and the main reference image [2].

 The disadvantage of this device is the lack of accuracy in determining the coordinates of the object.

 The purpose of the invention is to improve the accuracy of determining the coordinates of the object and providing the ability to determine the color of the object.

 This goal is achieved by the fact that in the device containing the imager of the current image (TI), as well as sequentially connected additional reference image (DEI), block correlation-extreme information processing (BKEOI), a shift mechanism and the main reference image (EI), the output is connected to a separate input of the correlation-extreme information processing unit, the output of the change mechanism being also connected to the second input of the additional reference image, electro-optical the first filter located between the optical output of the imager of the current image and the optical input of the additional reference image, the color switching unit, the output connected to the second input of the electro-optical filter, and BVKSS, the first input connected to the output of the correlation-extreme information processing unit, with two separate outputs of the selection block frame and horizontal sync pulses (BVKSS) are connected to the individual inputs of the color switching unit, and the second input BVKSS connected to the output of the shaper uschego image.

 The introduction of an electro-optical filter, a color switching unit, and BVCCS and their inclusion according to the indicated scheme allow us to divide the color spectrum into the main colors: red, green, blue, and define, besides the OM coordinates, such colors as red, green, blue, yellow, magenta, cyan and white. In addition, the use of a personal computer equipped with an interface for controlling the operation parameters of the device, indicating intermediate and final results, can significantly simplify device management, increase the visibility of the results, expand the range of external conditions to which the device can be adapted.

 In known devices that solve the problem of estimating OM coordinates based on the method of correlation-extreme comparison of TI and EI, color analysis of the compared images is not used, i.e. due to the lack of an electro-optical filter, a color switching unit, and BVKSS included according to the specified scheme. All this makes it impossible to determine the color of the OM or SE surface of the material. The presence of distinctive features not known in technical solutions allows us to conclude that the proposed solutions meet the conditions of an inventive step.

 In FIG. 1 shows a block diagram of a device. In FIG. Figure 2 shows the use of a part of the device, the Mayer-Eppler optical correlator. In FIG. 3 shows an optical diagram of one section of an electro-optical filter; FIG. 4 shows an embodiment (structural electrical circuit) of an electro-optical filter. In FIG. 5 shows an embodiment of a color switching unit. In FIG. Figure 6 shows plots of the voltages generated by the color switching unit. In FIG. 7 presents the implementation of cross sections of the cross-correlation function (VKF). In FIG. Figure 8 shows an embodiment of the main part of the correlation-extreme information processing unit — the VKF analyzer. In FIG. 9 shows a block diagram of the algorithm for the operation of the VKF analyzer. In FIG. 10 shows an embodiment of the BVKSS. In FIG. 11 shows voltage diagrams (video signal and clock pulses).

In FIG. 1 shows a structural diagram of the proposed device, where
1 - current image former (IIT);
2 - electro-optical filter (EOF);
3 - mask additional reference image (DEI);
4 - block correlation-extreme information processing (BKEOI);
5 - mechanism for changing the primary and secondary reference image (MS);
6 - mask of the main reference image (OEI);
7 - color switching unit (BPC);
8 - block allocation of personnel and horizontal sync pulses (BVKSS).

 From the block diagram shown in FIG. 1, it follows that the output of the Physicotechnical Institute 1 through series-connected EOF 2, DEI 3, BKEOI 4 is connected to MS 5; the output of the MS 5 is connected to the second input of the DEI 3 and the input of the OEI 6, the output of which is connected to the second input BKEOI 4; a separate output of BPC 7 is connected to a separate input of the EOF 2; two separate outputs BVKSS connected to the control outputs of BPC 7; the first input BVKSS connected to the output of the Physicotechnical Institute 1; the output of BKEOI 4 is connected to the second input of the BVKSS, a separate output of BKEOI 4 is the output of the device.

 The device operates as follows. On the screen of the Physicotechnical Institute 1, a color TI is reproduced, which is displayed either on the screen of a television monitor or on the screen of a color monitor containing a storage cathode ray tube (ZELT). A kit is used as a TI shaper: a color television camera and a color television monitor (receiver) (see sets [8, p. 313-315; 9, p. 5-11, 40-70]).

где J_R(ξ,η), J_G(ξ,η), J_B(ξ,η) - ВКФ, характеризующие соответственно красный R, зеленый G и синий B цвета сравниваемых изображений;
L'₁(x'₁,y'₁) - функция, описывающая ЭОФ 2;
L₁(x₁ ⁰,y₁ ⁰) - функция, описывающая ДЭИ 3;
F₁ ^R(x₁,y₁), F₁ ^G(x₁,y₁), F₁ ^B(x₁,y₁) - соответственно функции, описывающие ТИ красного, зеленого и синего цветов;
F₂ ^R(x₂,y₂), F₂ ^G(x₂,y₂), F₂ ^B(x₂,y₂) - функции, описывающие соответственно ЭИ красного, зеленого и синего цветов ОЭИ 6;
x₁o₁y₁, x'₁o'₁y'₁, x''₁o''₁y''₁, x₂o₂y₂ и ξoη - системы координат, связанные с соответствующими плоскостями;
K - коэффициент пропорциональности;
s - площадь коррелируемых изображений.Thus, when forming a TI, a television camera perceives a color image that is displayed on a color monitor screen. It should be noted that two signals are removed from the output of the Physicotechnical Institute 1: optical (i.e., the current image itself) and electric (video signal). The optical signal is fed to EOF 2, and the electric signal to BVKSS. Synchronously with the formation of the image on the screen of the Physicotechnical Institute 1, the corresponding field (for example, red) is included in the EOF 2. Thereby, a filter is formed corresponding to a certain wavelength of light (for example, red R color). It controls the operation of the BVKSS BKEOI 4, which generates the corresponding resolution pulse. Then, in the EOF 2, a filter of blue B and green G colors is also formed. The color image after passing through the EOF 2 is a monochrome image of red R, blue B or green G colors. Next, the image, after EOF 2, passes through the DEI 3, filtering the interference present on the TI. DEI 3 is a combination of transparent and opaque mask sections. The images of the manipulation object (or objects) on it are transparent, identical in shape and location to the TI, and the boundaries of the image of the objects are expanded compared to the TI by the amount of the permissible error in determining the offsets of the manipulated object in the working area of the robot. A monochrome TI filtered out from interference is introduced into the BKEOI 4, where OEI 6 is also introduced. OEI 6 is a mask, the transparent sections of which represent images of the manipulated object in the required reference position. Thus, the device generates and analyzes successively three color images of the VKF, described by the following expressions:

where J _R (ξ, η), J _G (ξ, η), J _B (ξ, η) - VKF, characterizing respectively the red R, green G and blue B colors of the compared images;
L ' ₁ (x' ₁ , y ' ₁ ) - a function that describes the EOF 2;
L ₁ (x ₁ ⁰ , y ₁ ⁰ ) is a function that describes the DEI 3;
F ₁ ^R (x ₁ , y ₁ ), F ₁ ^G (x ₁ , y ₁ ), F ₁ ^B (x ₁ , y ₁ ) - respectively, functions that describe TIs of red, green and blue colors;
F ₂ ^R (x ₂ , y ₂ ), F ₂ ^G (x ₂ , y ₂ ), F ₂ ^B (x ₂ , y ₂ ) - functions that describe, respectively, EI of red, green and blue colors OEI 6;
x ₁ o ₁ y ₁ , x ' ₁ o' ₁ y ' ₁ , x'' ₁ o'' ₁ y'' ₁ , x ₂ o ₂ y ₂ and ξoη are coordinate systems associated with the corresponding planes;
K is the coefficient of proportionality;
s is the area of correlated images.

Based on the analysis of expressions (1), the color of the manipulated object or a structural element of the material surface and its linear coordinates are estimated. The color of the OM or SE surface of the material is determined based on the implementation of the well-known logical rule [8, p. 343], from which it follows that the color C of the object can be described by the following color equation:
C = T ₁ [r _ij ] + T ₂ [g _ij ] + T ₃ [b _ij ],
where T ₁ , T ₂ , T ₃ are the coordinates of the color C;
r _ij , g _ij , b _ij are the unit quantities of the main corresponding colors R, G, B (their light fluxes, their brightness or the illumination created by them, taken as a unit). The color of OM or SE is determined by evaluating the combination of colors given in the table [8, p. 42].

 As can be seen from the table, the device can determine seven colors of OM or SE. To this end, BKEOI 4 on the basis of sequential fixing the amplitude of the VKF according to the table estimates the resulting OM color (for example, if the first signal is recorded for the entire cycle of the device, the resulting color will be red, and if the second and third signals are recorded, the resulting color will be blue) .

The operation for evaluating the color of an object is preceded by an operation for estimating the coordinates of an object. The coordinates ξ _i , η _{i of the} object are determined based on the analysis of the position of the main maximum of the CCF in the correlation plane. Further, the device’s operation cycle continues until it becomes necessary (change of the working area with OM or change of scale of the compared images) to change the reference images of DEI 3 and OEI 6 (see condition (3)). After changing the reference images of the DEI 3 and OEI 6, the work cycle resumes.

 In FIG. Figure 2 shows an embodiment of a part of the device, the Mayer-Eppler optical correlator, constructed in accordance with [1, p. 54, Fig. 2.24].

The correlator contains the following elements:
2 - electro-optical filter (EOF);
3 - additional reference image (DEI);
9 - current image (TI);
10 - matching part (undistorted) TI (STI);
11 - mismatching part (interference) TI (STI);
12 - image of OM or SE on DEI 3;
13 - reference image (EI);
14 - image of OM or SE on EI 13;
15 - plane registration VKF.

TI 9 is an element of the Physicotechnical Institute 1, and EI 12 and the registration plane VKF 15 are elements of the BKEOI 4. In the correlator, coordinate systems x ₁ o ₁ y ₁ , x ' ₁ o' ₁ y ' ₁ , x'' ₁ o'' _{1 are} introduced y '' ₁ , x ₂ o ₂ y ₂ and ξoη, respectively associated with the TI 9, EOF 2, DEI 3, EI 13 planes and the VKF 15 registration plane.

где M - линейный размер ТИ 9 относительно ДЭИ 3 (A, B - линейные размеры ТИ 9 и ДЭИ 3);
t, f - параметры коррелятора;
ΔX₁, ΔY₁, ΔX₂, ΔY₂ - координаты, связанные с ТИ 9 и ДЭИ 3;
Δξ, Δη - координаты главного максимума ВКФ в плоскости регистрации 15 (точка M').The parameters of the optical correlator are related by the following relationships:

where M is the linear size of TI 9 relative to DEI 3 (A, B are the linear dimensions of TI 9 and DEI 3);
t, f - correlator parameters;
ΔX ₁ , ΔY ₁ , ΔX ₂ , ΔY ₂ - coordinates associated with TI 9 and DEI 3;
Δξ, Δη - coordinates of the main maximum VKF in the registration plane 15 (point M ').

или

где x₁, y₁ - координаты центра тяжести любого элемента ТИ 9;
x₁ ⁰, y₁ ⁰ - координаты центра тяжести совпадающего элемента ТИ 9;
S⁰ - область существования совпадающего элемента;
ΔS⁰ - приращение области S⁰.The correlator works as follows. Color TI is projected on EOF 2, which passes a specific color image to DEI 3. Using DEI 3, the interference filtering 11 located on TI 9 is carried out. For this, DEI 3 is implemented as [1, p. 50]:

or

where x ₁ , y ₁ - coordinates of the center of gravity of any element of TI 9;
x ₁ ⁰ , y ₁ ⁰ - coordinates of the center of gravity of the coincident element TI 9;
S ⁰ is the region of existence of the matching element;
ΔS ⁰ is the increment of the region S ⁰ .

где δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{1}}$ - допустимая ошибка определения координат ОМ;
δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{2}}$ - смещение элементов на границе экрана ТИ 9, имеющего соответствующие размеры растра по горизонтали и вертикали A_р и B_р, при относительном угле ν_d разворота датчика ТИ (например, телекамеры) и конвейера;
δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{1}}$ , δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{2}}$ , δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{3}}$ определяются в масштабах изображений.As already indicated, the DEI 3 is a mask in which the images of the elements are transparent, identical in shape and location to the EI 13, and the boundaries of the image elements on the DEI 3 are expanded in comparison with the EI 3 by

where δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{1}}$ - permissible error in determining the coordinates of OM;
δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{2}}$ - the displacement of the elements on the border of the screen TI 9, having the corresponding horizontal and vertical raster sizes A _p and B _p , with a relative angle ν _{d of the} rotation of the TI sensor (for example, a camera) and the conveyor;
δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{1}}$ , δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{2}}$ , δ $\genfrac{}{}{0ex}{}{\text{0}}{\text{3}}$ defined on the scale of the images

 Thus, in the registration plane of the BKEOI 4, an optical image of the VKF will be formed, described by expression (1).

In FIG. 3 shows a block diagram of one section of the EOF 2 [4]. The EOF section 2 contains the following elements:
16, 19 - achromatic neutral polarizers;
17 - chromatic phase plate;
18, 20, 22 - electro-optical phase plates (EOFP);
21, 23 - chromatic polarizers;
24 - isotropic light filter.

 EOFP 18, 20 and 22, made of electro-optical ferroceramics, are divided into n independent sections, and the electrodes of each EOFP section 18 and 22 are interconnected (Fig. 4), forming the first inputs of the filter sections 24 [4]. The electrodes of the EOFP 20 are the second inputs of the sections of the filter 24. The first and second inputs of the i-th section of the EOF 2 are connected respectively to the nth and (n-1) -th outputs of BPC 7.

To obtain color-separated R (red), G (green) and B (blue) optical signals, EOF 2 is placed between the screen of the Physicotechnical Institute 1 and DEI 3. Each of the n sections of EOF 2 can pass red, green, and blue, depending on a combination of voltages applied to the electrodes of the electro-optical elements of the sections. In EOF 2, the transmission axes of the neutral polarizers 16 and 19 are oriented in parallel. The axis of orientation of the molecules of the polymer chromatic phase plate 17 is oriented at an angle of 45 ^o to the axes of the neutral polarizers 16 and 19. EOFPs 18, 20 and 22 made of electro-optical ferroceramics with transverse electro-optical effect are oriented in such a way that the direction of the control electric field is 45 ^o axes of neutral polarizers 16 and 19. Neutral polarizers 16 and 19 polarize unpolarized light.

EOFP 18, 20 and 22 when applying half-wave control voltages to their electrodes, they rotate the light polarization vector by 90 ^o . The chromatic phase plate 17, placed between parallel neutral polarizers 16 and 19, transmits red and blue, and with the perpendicular position of the axes of the neutral polarizers 16 and 19 transmits green light. The chromatic polarizer 21, when the direction of the light polarization vector coincides with the axis of maximum transmission, transmits all spectral components of light, and when the direction of the light polarization vector is perpendicular, it transmits only red light. The chromatic polarizer 23, when the direction of the polarization vector of light coincides with the axis of maximum transmission, transmits all spectral components of light, and with the perpendicular direction of the polarization vector transmits only blue light. An isotropic filter 24 suppresses the infrared region of the spectrum.

 In the initial state, i.e. in the absence of voltage, each section of the EOF 2 transmits light in the blue region of the spectrum. In order for any section of the EOF 2 to transmit red light, it is necessary to apply a half-wave voltage to the electrodes of the corresponding section of the EOF 20. When applying half-wave control voltages to the electrodes of any section of the EOFP 18, 20 and 22 of the EOF 2 section, green light is transmitted. To move the border of the color change of the EOF 2 after the deployment line of the raster of the transmitting television tube BKEOI 4, BPC 7 is used.

The block diagram of the BOC 7 is shown in FIG. 5. BOC 7 contains the following elements [4]:
25 - frequency divider into three;
26, 27 - triggers;
28 - shift register for two digits;
29, 30 - shift register for n bits;
31.1-31.n - pulse amplifiers;
32.1-32.n - pulse amplifiers.

 BOC 7 works as follows. Two impulse sequences with a frequency F (Fig. 6a) and nF (Fig. 6b), where F is the frequency of the television scan fields, n is the number of sections of the EOF 2, are received at the inputs of BPC 7 from the BVCCS 8. Pulses with the frequency of the fields are fed to the input of the divider frequency of three 25, the output of which is formed by a sequence of pulses with a frequency of F / 3 (Fig. 6B). From the output of the frequency divider by three 25 pulses are fed to the first input of the shift register into two bits 28. The second input of the shift register 28 receives pulses from the BVCCS with a frequency F. At the outputs of the shift register 28, sequences of pulses of frequency F / 3 are shifted relative to the sequence of pulses applied to the first input of the shift register 28 by a time T equal to the period of the field frequency at the first output of the shift register 28 (Fig. 6d) and by the time 2T at the second output of the shift register 28 (Fig. 6e).

 The pulses from the first and second outputs of the shift register 28 are respectively supplied to the first and second inputs of the trigger 26, forming at the output of the trigger 26 the pulse sequence shown in FIG. 6th. From the output of the frequency divider by three 25 pulses are supplied to the first input of the trigger 27. The pulses from the second output of the shift register 28 are supplied to the second input of the trigger 27. At the output of the trigger 27, the pulse sequence shown in FIG. 6g. From the output of the trigger 26, the pulses are fed to the first input of the shift register for n bits 29, to the second input of which pulses from the BVCCS with a frequency of nF are applied (Fig. 6b). At the first output of the shift register 29, the pulse train shown in FIG. 6th. A sequence of pulses is generated at the second output of the shift register 29 (Fig. 6h), shifted relative to the sequence of pulses at the first output of the shift register 29 by a time t equal to the period of pulses arriving at the second input of the shift register 29. At the n-th output of the shift register 29 the pulse sequence (Fig. 6i) shifted relative to the pulse sequence at the first output of the shift register 29 by the time (n-1) t.

 From the output of the trigger 27, the pulses are fed to the first input of the shift register for n bits 30, to the second input of which pulses from the BVCCS with a frequency of nF are fed. At the first output of the shift register 30, the pulse train shown in FIG. 6g, a pulse train is formed at the second output of the shift register 30 (Fig. 6k), shifted relative to the pulse train at the first output of the shift register 30 by time t. At the nth output of the shift register 30, a pulse sequence is formed (Fig. 6l), shifted relative to the pulse sequence at the first output of the shift register 30 by the time (n-1) t.

 From the outputs of the shift registers 29 and 30, the signals are respectively transmitted to pulse amplifiers 31.1-31.n and 32.1-32.n, where the voltage of the pulses is amplified to a value equal to the half-wave voltage of the EOFP 18, 20, and 22.

 From the outputs of the pulse amplifiers 31.1-31.n, the pulses are respectively supplied to the inputs of sections 2.1-2.n of the EOFP 18 and 22 of the electro-optical filter 2, and from the outputs of the pulse amplifiers 32.1-32.n, the pulses are respectively sent to the inputs of the sections 2.1-2.n of the EOF 20 electro-optical filter 2.

Therefore, at time t _cn , no voltage is applied to all sections of the EOFP 18, 20 and 22, and all sections of the electro-optical filter 2 remain blue, that is, they transmit blue. At time t _{k1, a} half-wave voltage is applied only to section 2.1 of the EOFP 20 and section 2.1 of the electro-optical filter 2 turns red, and the remaining sections (2.n-1) turn blue. At time t _k2, half-wave voltages are applied to sections 2.1 and 2.2 of the EOF 20. Therefore, sections 2.1 and 2.2 become red. At time t _kn, voltages are applied to all sections 2.n of the EOFP 20 and the entire electro-optical filter 2 turns red. At time t _P1 voltage is supplied to all sections, EOFP 20 and section 2.1 EOFP 18 and 22 and section 2.1 electrooptical filter 2 becomes green, and the remaining section 2.n - red. At time _tz2, voltage is applied to all sections of EOFP 20 and to sections 2.1 and 2.2 of EOFP 18 and 22. Consequently, sections 2.1 and 2.2 of electro-optical filter 2 become green. At time _tzn, voltage is applied to all sections of EOFP 18, 20 and 22, and the entire electro-optical filter 2 turns green. At time t _c1, voltage is applied to all EOFs 18, 20 and 22, except for section 2.1, which turns blue, and the remaining sections remain green, etc.

одного или нескольких цветных ОМ.In FIG. Figure 7 shows realizations of the VKF section along the oξ axis in the plane of registration of the VKF. Here in FIG. 7a shows a section of a VKF formed by three monochrome TI and EI. This corresponds to the case when the OM on the TI is represented by a color TI painted with a complex color. In FIG. 7b shows a section of the VKF formed by three monochrome TI and EI. This corresponds to the case when the TI is represented by two colored OM or SE (one in red R color and the second in blue (B + G) color). From the above implementations, it can be seen that color and coordinates can be estimated

one or more colored OM.

In FIG. Figure 8 shows an embodiment of the main part of BKEOI 4, an analyzer of correlation functions. This analyzer, implemented according to [1, p. 256; 6], contains the following elements:
33 - video signal sensor;
34, 35, 36 - blocks for the selection of read signals (BSSS);
37 - the first video amplifier;
38 - analog-to-digital Converter (ADC);
39 - the first register;
40 - memory block (PSU);
41 - multiplexer;
42 - second register;
43 - digital-to-analog converter (DAC);
44 - second video amplifier;
45 - pulse generator;
46 - block forming the address (BFA);
47 - shaper recording signals (FSZ);
48 - multiplexer;
49 - driver control signals (FUS);
50 - decoder;
51 - the first register;
52 - the second register;
53 - communication unit;
54 - monitor;
55 - computer;
56, 61 - interface (for example, standard I2);
57 - analog block;
58 - information input;
59, 60, 62 - information outputs.

 The memory blocks form three identical channels of the corresponding color: red, blue and green, so in the future we will consider the operation of only one of them.

 The analyzer works as follows. The video signal from the output of the sensor 33 is fed to a video amplifier 37 and, after amplification and digitization by means of the ADC 38, is fed to the input of the register 39, where it is entered by pulses from the output of the generator 45. The image frame is recorded in the power supply unit 40 by a command received from the computer 55 via the communication unit 53 The decoder 50 decodes this command and generates a signal to the state register 52, which waits after this arrival of a frame clock, and then generates a control signal to the FSZ 47 and register 39 of block 34. According to this signal, register 39 converts the last The output information coming from the ADC 38 is in parallel, after which the code is written to the PSU 40 using the recording signal from the FSZ 47. During the conversion, the FSF 49 supplies the BP 40 with the signals necessary to select the address for recording, and the BPA 46 through the multiplexer 48 provides sequential switching of bits to the address inputs. In this case, the joint operation of the register 39 and BFA 46 is synchronized by the pulse generator 45.

 With the end of the half-frame (with the arrival of the next frame sync pulse to register 52), register 52 removes the enable signal from register 39 and FSZ 47 and the device goes into control mode.

 In the control mode, information is read from the PSU 40 in a parallel code, converted using the register 42 into a serial one and fed to the DAC 43, the output of which appears the corresponding voltage level, which is amplified by the video amplifier 44 and fed to the monitor 54 through the output 59.

 When reading data from the PSU 40 to the output of the device, the decoder 50 outputs signals to the FSF 49 and the communication unit 53. According to this signal, the FSF 49 after the end of the next cycle of access to the PSU 40 sets a command to the multiplexer 48, by which an address is supplied to the address inputs of the PSU 40, pre-recorded in register 51.

 Then BP 40 control signals are generated - in the same way as in the control mode. As a result of this, the data read from this PSU 40, through the multiplexer 41 are output 60 and then to block 53. The multiplexer 41 selects, depending on the two least significant bits of the address, the corresponding group of microchips of the memory block 40.

 With the appearance of the signal at the output of the PSU 40 FUS 49 gives a signal to the decoder 50, which generates a signal for receiving information, by which the computer 55 proceeds to receive information. At the same time, a signal is taken through which register 51 is connected to BP 40 through multiplexer 48, and the correlation function analyzer goes into control mode.

 When writing data to the BP 40 and computer 55, the operation of the device is similar to reading, except that the decoder 50 provides a signal to the communication unit 53, connecting the output of the multiplexer 41 to the computer bus 55. In addition, the decoder 50 gives a signal to the FSZ 47, according to which to the group of microcircuits selected by the two least significant bits of the column address received at FSZ 47, a write signal is issued.

 Data in the BP 40 comes from the communication unit 53 through the register 39. The address in the register 51 is entered previously from the computer channel 55.

 Thus, the correlation function analyzer allows you to read the color image of the VKF, write from the computer the color image of the VKF in the memory and monitor the information recorded in the memory unit on the color monitor screen. Note here that the image to be read may be an image with a size of 256 x 256 elements containing 16 gradations of brightness.

 After that, the computer 55 (for example, "Electronics-60" or IBM PC XT, AT) begins to analyze color VKF. As a result, the color and coordinates of objects located in the working area of the robot are determined. In the analysis of VKF, the amplitude, width, steepness (or only the amplitude) and the position of the main maximum of the VKF in the plane of its registration can be studied 15 [1, 3].

After the analysis of the VKF using a computer is determined, the need to change the frame of the EI by analyzing the expression
Δξ _s -Δξ _i ≤ 0, (3)
where Δξ _s , Δξ _i are respectively given and current coordinates of the main maximum of the VKF in the plane of its registration. If expression (3) is not satisfied (i.e., Δξ _s > Δξ _i ), then a signal is transmitted via the interface 56 to the analog block 57, which converts the digital signal into an analog one. The interface is an I2 device [7]. Block 57 contains a DAC (chip 1118PA1, included in the standard scheme). Through the interface 61 receives information about the color and coordinates.

The generalized algorithm of the device operation is shown in FIG. nine:
63 - beginning;
64 - formation of an enable signal;
65 - input image VKF;
66 - calculation of coordinates;
67 - binarization (determination of the coefficients r _ij , g _ij , b _ij );
68 - memorization in RAM of the reference coefficients r _ij ^e , g _ij ^e , b _ij ^e ;
69 - color definition;
70 - verification of condition (3): Δξ _s -Δξ _i ≤ 0;
71 - record in RAM a predetermined value Δξ _s ;
72 - the formation of teams change personnel OEI and DEI;
73 - output of the result;
74 - the end.

 After executing the "Start" command, which conducts the description of the registers (block 63), the enable signal (block 64) of the BVCCS is supplied and the VKF image is input into the BP 40 of the VKF analyzer (block 65) shown in FIG. 8. The input image of the VKF can be displayed (displayed) on the monitor (video monitoring device).

характеризуют уровни красного, зеленого и синего цветов соответственно для каждого элемента z_ij массива z. После введения этих трех изображений согласно (1) определяются координаты Δξ_i,Δη_i наибольшего значения ВКФ (блок 66). Хотя координаты Δξ_i,Δη_i могут определяться для каждого цветового поля, но в данной программе, с целью упрощения, координаты Δξ_i,Δη_i/ оцениваются только по массиву R (красной компоненте). Такое допущение всегда правомерно, если объект манипулирования имеет простую цветовую окраску (например, объект желтого или пурпурного цвета). После введения этих трех изображений запоминаются и определяются уровни (амплитуды) r_ij, g_ij, b_ij (блок 67) изображений. Запомненные коэффициенты r_ij, g_ij, b_ij сравниваются (блок 69) с эталонными (табличными) r_ij ^э, g_ij ^э, b_ij ^э (блок 68). После этого принимается решение (например, при максимальном совпадении вычисленных и эталонных коэффициентов (см. таблицу)) о цвете изображения ВКФ и, следовательно, цвете объекта, находящегося в рабочей зоне.The image entered is a collection of arrays (block 68) R, G, B, the elements of which

characterize the levels of red, green and blue, respectively, for each element z _{ij of} array z. After the introduction of these three images, according to (1), the coordinates Δξ _i , Δη _{i of the} highest VKF value are determined (block 66). Although the coordinates Δξ _i , Δη _i can be determined for each color field, in this program, for the sake of simplicity, the coordinates Δξ _i , Δη _{i /} are estimated only using the R array (red component). Such an assumption is always valid if the object of manipulation has a simple color (for example, an object of yellow or purple). After the introduction of these three images, the levels (amplitudes) r _ij , g _ij , b _ij (block 67) of the images are stored and determined. The stored coefficients r _ij , g _ij , b _{ij are} compared (block 69) with the reference (tabular) r _ij ^e , g _ij ^e , b _ij ^e (block 68). After that, a decision is made (for example, at the maximum coincidence of the calculated and reference coefficients (see table)) about the color of the VKF image and, therefore, the color of the object in the working area.

Using block 70, condition (3) is checked by comparing the stored value Δξ _s (block 71) and the current value Δξ _i . Based on this analysis, a decision is made (block 72) to change the OEI 6 and DEI 3 frames. When these frames are changed, the work cycle stops and then resumes again from the beginning. After the operation of outputting the result (block 73), the program ends (block 74).

 Listing (text) of the main part of the program that implements the described algorithm for working with the VKF analysis is given below. The work program of the OCEI 4 is written in C language. Briefly explain the program. The program begins with a description of the registers (external devices) of the correlation function analyzer and the supply of an enable signal to the control unit. Then, the components of the frame of the VKF image are input. After that, the coordinates of the points of maximum brightness are determined and the binarization operation is performed. If necessary, the reference image is changed to stop the cycle of the device.

#include <dos.h>
#include <conio.h>
#include <stdio.h>
/ * description of registers of the analyzer of correlation functions * /
#define RSB 164000
#define RST 164002
#define RS 164004
#define RDR 164012
#define RDG 164014
#define RDB 164016
#define RC 167700
void main (void)
{
int i, j, k, r_max = 0, ksi, nu, i_cont, j_cont;
int i_kor, j_kor, r, g, b, color;
char red, green, blue;
char pr, pg, pb;
/ * enable signal to the control unit * /
BEG: outportb (RC, 1);
/ * input of the components of the frame of the image VKF * /
STR: outportb (RS, 9);
outportb (RS, 10);
outportb (RS, 11);
/ * determination of coordinates of points of maximum brightness * /
for (i = 0; i <= 255; i ++) {
for (j = 0; j <= 255; j ++) {
outportb (RSB, i);
outportb (RST, j);
k = inportb (RDR);
if (k> r_max) {
r_max = b;
i_kor = i;
j_kor = j;
}
}
}
/ * "binarization" * /
pr = 7; pg = 4; pb = 9;
for (i = 0; i <= 255; i ++) {
for (j = 0; j <= 255; j ++) {
outportb (RSB, i);
outportb (RST, j);
r = inportb (RDR);
g = inportb (RDG);
b = inportb (RDB);
if (r> pr) outport (RDR, 15);
else outport (RDR, 0);
if (g> pg) outport (RDG, 15);
else outport (RDG, 0);
if (b> pb) outport (RDB, 15);
else outport (RDB, 0);
}
}
/ * color of the maximum intensity point * /
outportb (RSB, i_kor);
outportb (RST, j_kor);
if ((r = inportb (RDR)) == 15) red = 100;
else red = 0;
if ((g = inportb (RDG)) == 15) green = 10;
else green = 0;
if ((b = inportb (RDG)) == 15) blue = 1;
else blue = 0;
color = red + green + blue;
switch (color) {
case 111: printf ("white");break;
case 110: printf ("yellow");break;
case 11: printf ("cyan");break;
case 10: printf ("green");break;
case 101: printf (magenta); break;
case 100: printf ("red");break;
case 1: printf ("blue");break;
case 0: printf ("black");break;
}
/ * re-enter * /
i_cont = 100;
j_cont = 100;
if (i_cont <= i_kor) {ksi = i_kor; goto STR;}
if (j_cont <= j_kor) {nu = j_kor; goto STR;}
goto CNG;
/ * change of standard * /
CNG: printf ("row coordinate =% d, column =% d", ksi, nu);
outportb (RC, 2);
goto BEG;
}
In FIG. 10 shows the electrical circuit BVKSS. The input from the BVKSS 8 receives a video signal from the camera of the Physicotechnical Institute 1. The BVKSS generates two pulse sequences with a frequency of F and nF, where F is the frequency of the television scan fields, n is the number determined by the number of lines of the camera. These signals are necessary for the operation of the BOC 7. The BVCCS extracts frame (Fig. 11b) and lowercase (Fig. 11c) clock pulses from the video signal (Fig. 11a). In FIG. 11 the following notation is used: U - voltage; t is time. These clock pulses arrive at chip 75 (155LI1); frame and line pulses are supplied to the corresponding inputs of BPC 7. The BVCCS is built on chip 76 (174XA11), which is used in standard power-on.

 MS 5 is implemented and functions similarly to that described in [10].

где N₁ - число правильных ответов (правильного распознавания);
N₂ - общее число испытаний (общее число объектов манипулирования).Compared with the known, the proposed device has higher accuracy characteristics. Show it. Suppose, for example, the task of sorting by color the OM entering the work area. Moreover, OM can be of seven colors (see table). If OMs have the same shape but different color, then the probability P of the correct color recognition is determined by the well-known expression:

where N ₁ is the number of correct answers (correct recognition);
N ₂ - the total number of tests (total number of objects manipulated).

а для предлагаемого вероятность правильного распознавания будет

Отметим здесь, что величина P₁ для известного устройства соответствует единичному случаю правильного распознавания, когда ОМ представлен объектом белого цвета.Then, for a known device, the probability of correct recognition will be

and for the proposed probability of correct recognition will be

We note here that the value of P ₁ for the known device corresponds to a single case of correct recognition when the OM is represented by an object of white color.

лучшими точностными характеристиками по сравнению с известным устройством.Thus, the proposed device has 7 times

better accuracy characteristics than a known device.

Literature
1. Korikov AM, Syryamkin VI, Titov BC Correlation visual systems of robots. - Tomsk: Radio and communications, Tomsk. Department, 1990 .-- 264 p.

2. A. p. N 532283, M. Cl. ² G 01 C 21/14, G 01 S 7/22. A device for determining the coordinates of an object // Angels M.P., Karpov A.G., Katyshev V.A. et al. 06.23.75, application N 2147734/23 (prototype).

 3. Correlation-extreme video sensor systems for robots // Andreev Yu. A., Beloglazov NN, Korikov AM, Syryamkin VI, Tarasenko VP - Tomsk: Tomsk Publishing House. un-that. - 240 p.

 4. A.S. N 1249718, M. Cl. H 04 N 5/225. Single-tube color television camera // Tsaplin M.N., Dadeshidze V.V., Pomberg M.G., Sokolov V.A. and other publ. 08/07/86, Bull. N 29.

5. A. p. N 1352668, M. Cl. ⁴ H 04 N 9/09. Color television camera // Ukhanov S.P., Only V.V. Publ. 11/15/87, Bull. N 42.

6. Device for reading and displaying images of objects // Nechunaev P.I., Syryamkin V.I., Titov V.S. and other Application N 4615286 / 24-24 (150511), decl. 10.25.88, M.C. ⁴ G 06 K 9/00, 11/00. The positive decision of 04/18/89.

 7. Zakharov IV. Maintenance and operation of the microcomputer "Electronics-60M". - M.: Mechanical Engineering, 1989 .-- 192 p. (p. 75).

 8. Kryzhanovsky V.D., Kostykov Yu.V. Television color and black and white. - M.: Communication, 1980 .-- 336 p.

 9. Only VV, Ozhigin A.F., Kharitonov Yu.A. Portable color television cameras. - M .: Radio and communications, 1984. - 104 p.

 10. A.S. N 1823773, M. Cl. H 04 N 7/18. A device for determining the coordinates of an object // Angels MP, Kondychekov GB, Korikov AM et al. 07/30/90, application N 4855747/09.

Claims (1)

 A device for determining the coordinates and color of an object, containing the imager of the current image, as well as sequentially connected mask of the additional reference image, a block of correlation-extreme information processing, a mechanism for changing masks of the main and additional reference images, and a mask of the main reference image, the output connected to a separate input of the correlation block -extreme information processing, and the output of the mechanism for changing the masks of the primary and secondary reference images also connected to the second input of the mask of the additional reference image, characterized in that it additionally includes an electro-optical filter connected between the optical output of the imager of the current image and the optical input of the mask of the additional reference image, a color switching unit, an output connected to the second input of the electro-optical filter, and the unit allocation of personnel and horizontal sync pulses, the first input connected to the output of the block correlation-extreme information processing, pr why two separate outputs of the block for selecting frame and horizontal sync pulses are connected to separate inputs of the color switching unit, and the second input of the block for selecting frame and horizontal sync pulses is connected to the output of the current image shaper.

Images