CN101872423A - A method for tracking moving objects on a production line - Google Patents

Abstract

The invention discloses a method for tracking a moving target on a production line, which belongs to the field of automation equipment. A visual sensor collects images of moving targets and transmits the images to a computer; a servo motor controls the running speed of a conveyor belt to obtain displacement information of the conveyor belt; the computer receives the image to identify the position information of the moving target; the computer obtains the effective position of the moving target based on the image, the position information, the first decision-making condition, and the second decision-making condition, excluding the situation that the moving target is omitted or repeatedly identified information; the computer obtains the position coordinates of the moving target according to the effective position information and the displacement information, and saves the position coordinates in the database of the captured moving target; The position coordinates of the moving target are sequentially extracted from the database, and a grabbing operation is performed. This method can meet the requirements of high-speed and real-time grasping of scattered moving objects by parallel manipulators.

Description

A method for tracking moving objects on a production line

technical field

The invention relates to the field of automation equipment, in particular to a method for tracking moving objects on a production line.

Background technique

In the modern industrial production process, a large number of industrial robots work in a structured environment with known parameters, relying on precise position settings to grab and place workpieces, which improves production efficiency and product quality. To a certain extent It meets the needs of industrial production and people's life. At the same time, they also have great limitations, because most robots require the workpiece to appear in a fixed position with precise orientation. When this requirement cannot be met, it often leads to failure or interruption of the production process, and also That is to say, these robots do not have the ability to perceive changes in the external environment. At present, even the most intelligent robot in the world, its ability to adapt to changes in the external environment is very limited, and it is still far from the goal expected by people, which greatly limits the promotion and application of robots, and one of them The important reason is that robots lack human-like perception capabilities. In order to solve this problem, researchers began to install various sensors for industrial robots, one of the more important ones is the vision sensor. Vision is an important means for human beings to observe and perceive the world. According to statistics, about 80% of the information that humans obtain from the external world is obtained by vision. This not only shows that the amount of visual information is huge, but also shows that humans have a high utilization rate of visual information, and at the same time reflects the importance of human visual function. With the development of information technology, it has been a dream of human beings for many years to give computers, robots or other intelligent machines the function of human vision. In the process of modern industrial automation production, machine vision technology is becoming a key technology to improve production efficiency and ensure product quality, such as automatic detection of mechanical parts, intelligent robot control and automatic monitoring of production lines, automatic tracking and identification of moving targets, etc. .

At present, the commonly used moving target tracking method is mainly to extract the target features in the video image and establish a feature model. The typical method is to use the color space characteristics to track the target in real time, and use the edge contour of the target to separate the target from the background for tracking. There are also many methods for modeling time dimension characteristics. Gray value is often used as the feature of target tracking, such as frame difference method of image sequence, Gaussian mixture model, adaptive filtering method, hidden Markov model, etc. The tracking of the moving target can be realized through the above method and model.

In the process of realizing the present invention, the inventor finds that the above-mentioned prior art has at least the following disadvantages and deficiencies:

The above methods and models have high computational complexity, and it is difficult to meet the high-speed and real-time requirements of the production line robot to grasp the moving target, and cannot meet the needs of practical applications.

Contents of the invention

In order to meet the high-speed real-time requirements of the production line robot to grab the moving target, the present invention provides a method for tracking the moving target on the production line. The method includes the following steps:

(1) The visual sensor collects the image of the moving target, and transmits the image to the computer;

(2) The servo motor controls the running speed of the conveyor belt and obtains the displacement information of the conveyor belt;

(3) The computer receives the image in step (1), and identifies the position information of the moving target;

(4) According to the image in step (1), the position information in step (3), the first decision-making condition, and the second decision-making condition, the computer excludes the situation that the moving target is missed or repeatedly identified, and obtains the Effective location information of the moving target;

(5) The computer obtains the position coordinates of the moving object according to the effective position information in step (4) and the displacement information in step (2), and saves the position coordinates in the captured moving object in the database;

(6) The parallel manipulator sequentially extracts the position coordinates of the moving target from the database of the grasped moving target in step (5), and performs a grabbing operation.

The first decision-making condition in step (4) is specifically:

为运动目标在x轴方向的最大长度。 Among them, ds is the number of pixels moved by the conveyor belt, M is the length of the field of view, the coordinates of the moving target in the image are two-dimensional coordinates, x is the abscissa of the moving target, y is the vertical coordinate of the moving target, and the direction of the x coordinate is the same as that of the conveyor belt The forward direction is the same, the y direction is consistent with the width direction of the field of view,

is the maximum length of the moving target in the x-axis direction.

The second decision-making condition in step (4) is specifically:

The set of coordinates of multiple moving targets obtained in the first frame of images in two adjacent frames is A={( _xi , y _i )|0< _xi , y _i <M, i∈N}, the second The coordinate set of multiple moving targets obtained in the frame image is B={(x _j , y _j )|0<x _j , y _j <M, j∈N}. The position information of the moving target is compared one by one;

When x _{j -xi} = ds and y _j -y _i = 0, it means that the jth moving object in set B is the repetition of the ith moving object in set A, and the position information corresponding to the jth moving object in set B should be discarded;

When x _{j -xi} = ds and y _j -y _i ≠0, it means that the jth moving object in set B is located on the same vertical line as the i-th moving object in set A, and the jth moving object in set B corresponds to The location information of should be kept;

When x _j -xi _≠ ds, it means that the j-th moving object in set B is not the same moving object as the i-th moving object in set A, and the position information corresponding to the j-th moving object in set B should be retained;

Among them, ds is the number of pixels moved by the conveyor belt, M is the length of the field of view, and N is the width of the field of view. The coordinates of the moving object in the image are two-dimensional coordinates, x is the abscissa of the moving object, and y is the vertical direction of the moving object coordinate.

The computer in step (5) obtains the position coordinates of the moving target according to the effective position information in step (4) and the displacement information in step (2), specifically:

Merge the position information of the remaining moving objects in the coordinate set of the moving objects obtained in the second frame image into the set T, T={(x _k , y _k )|0<x _k , y _k <M , k∈N};

Add the displacement information of the conveyor belt to the corresponding points in the set T, and update the set T to T={(x _k , y _k , c _k )|0<x _k , y _k <M, k∈N};

Obtain the position coordinates of the moving target during the operation of the conveyor belt;

$\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)<span\; class="notranslate">\; =</span>\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; dc</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}\end{array}\right)$

Among them, the set T is recorded as the position information set of all moving objects correctly identified, M is the length of the field of view, and N is the width of the field of view. The coordinates of the moving object in the image are two-dimensional coordinates, and x is the abscissa of the moving object , y is the ordinate of the moving target, c _k is the pulse number of the conveyor belt motor at the moment of image acquisition at this point, (x, y) ^T is the position coordinate of the moving target on the conveyor belt when the pulse number of the conveyor belt motor is c, dc is the distance traveled by the unit pulse conveyor belt.

The displacement information in the step (2) is specifically:

By reading the pulse value in the servo motor, the current displacement information of the conveyor belt is obtained.

A two-way feedback control is formed between the servo motor and the vision sensor.

The beneficial effects of the technical solution provided by the invention are:

The method collects the image information of the moving target at a fixed position through the visual sensor, obtains the image coordinates of the moving target through calculation, and combines the displacement information provided by the conveyor belt, the first decision-making condition, and the second decision-making condition to track the position coordinates of the moving target. Pass it to the parallel manipulator to provide positioning information for the parallel manipulator to perform reliable and fast grabbing operations, which meets the requirements of the automated production line; avoids the situation that the moving target is repeatedly identified or is missed; and through the frequency of servo motor and visual sensor The two-way feedback control between the parallel manipulators meets the requirements of high-speed and real-time grasping of scattered moving targets.

Description of drawings

Fig. 1 is the automatic production line layout diagram provided by the present invention;

Fig. 2 is the flow chart of the moving object tracking method on the production line provided by the present invention;

Fig. 3 is the schematic diagram that the moving object provided by the present invention appears in the previous frame image of two adjacent frame images;

Fig. 4 is the schematic diagram that the moving target provided by the present invention appears in the current frame image of two adjacent frame images;

Fig. 5 is a schematic diagram of the position of the moving target provided by the present invention;

Fig. 6 is a schematic diagram of the first frame image of continuous shooting provided by the present invention;

Fig. 7 is a schematic diagram of the second frame image of continuous shooting provided by the present invention;

Fig. 8 is a schematic diagram of the third frame of images continuously shot provided by the present invention.

In the accompanying drawings, the list of parts represented by each label is as follows:

1: Visual sensor, 2: Parallel manipulator, 3: Conveyor belt, 4: Several moving targets, 5: Servo motor.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the implementation manner of the present invention will be further described in detail below in conjunction with the accompanying drawings.

In order to meet the high-speed and real-time requirements of the production line robot to grab the moving target, the embodiment of the present invention provides a method for tracking the moving target on the production line, see Figure 1 and Figure 2, the content of the method is as follows:

Figure 1 includes: visual sensor 1, parallel manipulator 2, conveyor belt 3, several moving objects 4, and servo motor 5. The visual sensor 1 and the parallel manipulator 2 are arranged above the conveyor belt 3 , and several moving objects 4 are scattered on the conveyor belt 3 and move in the direction of the parallel manipulator 2 along with the conveyor belt 3 .

Two-way feedback control is formed between the servo motor 5 and the visual sensor 1, specifically: when the visual sensor 1 captures more moving objects 4 at the current moment, in the subsequent steps, there will be problems caused by recognizing more moving objects 4. Excessive time consumption will cause the parallel manipulator 2 to grab and respond untimely. In order to avoid this phenomenon, the operating speed of the servo motor 5 needs to be reduced. The visual sensor 1 sends a signal to the servo motor 5, and the servo motor 5 controls the conveyor belt after receiving the signal 3 Reduce the running speed; after the servo motor 5 controls the conveyor belt 3 to reduce the running speed, the feedback signal is sent to the visual sensor 1, and the visual sensor 1 reduces the shooting frequency; on the contrary, when the moving target 4 at the current moment captured by the visual sensor 1 is less, it means At this time, the running speed of the conveyor belt 3 is relatively slow, and the running speed needs to be increased. The visual sensor 1 sends a signal to the servo motor 5, and the servo motor 5 controls the conveyor belt 3 to increase the running speed; after the servo motor 5 controls the conveyor belt 3 to increase the running speed, it feeds back to the visual sensor 1. Visual sensor 1 increases the shooting frequency.

101: The visual sensor collects the image of the moving target, and transmits the image to the computer;

When the moving target passes under the field of view of the visual sensor, the visual sensor collects the image of the moving target and transmits the collected image of the moving target to the computer.

Wherein, in order to obtain the image of the moving target conveniently and quickly, preferably, the vision sensor in the embodiment of the present invention is based on a CCD (Charge-coupled Device, charge-coupled device) chip, and during specific implementation, it can also be based on other A type of chip, for example: a visual sensor based on a CMOS (Complementary Metal Oxide Semiconductor, Complementary Metal Oxide Semiconductor) chip, which is not limited in the embodiment of the present invention.

102: The servo motor controls the running speed of the conveyor belt, and obtains the displacement information of the conveyor belt;

Specifically, the moving target moves with the conveyor belt to the grasping space direction of the parallel manipulator. By reading the pulse value in the servo motor, the current displacement information of the conveyor belt, that is, the distance advanced, can be obtained.

103: The computer receives the image transmitted by the visual sensor in step 101, and identifies the position information of the moving target;

104: According to the image of the moving object collected in step 101, the location information in step 103, the first decision-making condition, and the second decision-making condition, the computer excludes the situation that the moving object is missed or repeatedly identified, and obtains the effective location information of the moving object ;

会避免遗漏识别目标，其中，

表示运动目标在x轴方向的最大长度。Because the moving objects are scattered on the conveyor belt, there is no uniform rule, so that in the single frame image collected by the visual sensor, only a part of some moving objects may enter the camera field of view of the visual sensor, so that the computer cannot recognize the incomplete movement. The target is identified, resulting in omissions. In order to avoid the above-mentioned situation, the embodiment of the present invention provides a first decision-making condition. The first decision-making condition is specifically: after the visual sensor is fixed, its field of view is also determined. Set the field of view The aspect ratio of the image is M×N (in pixels), where M is the length and N is the width. The coordinates of the moving target in the image are two-dimensional coordinates, x is the abscissa of the moving target, and y is the vertical coordinate of the moving target , the x coordinate direction is consistent with the advancing direction of the conveyor belt, and the y direction is consistent with the width direction of the field of view. Every time the conveyor belt moves ds pixels, the camera in the visual sensor 1 takes a picture.

will avoid missing recognition targets, where,

Indicates the maximum length of the moving target in the x-axis direction.

Specifically, since the moving object on the conveyor belt moves forward together with the conveyor belt, the obtained position information of the moving object only differs by ds pixels in the x-coordinate direction, and there is no change in the y-coordinate direction. See Fig. 3 and Fig. 4. In Fig. 3, there are 4 moving targets in the field of view M×N, including 3 complete moving targets (No. 2 moving target, No. 3 moving target, and No. 4 moving target) and 1 different The complete moving target (moving target No. 1), Fig. 4 shows that after the conveyor belt has moved forward for a distance ds, there are 3 moving targets in the field of view M×N, including 2 complete moving targets (moving target No. 3 and moving target 4 No. moving target) and 1 incomplete moving target (No. 2 moving target). By analyzing Figure 3 and Figure 4, it can be known that the No. 3 moving target and the No. 4 moving target completely appear in the two frames of images. After analysis, the following conclusions are obtained:

时，两帧图像中可能会有重复且完整的运动目标；1. When

When , there may be repeated and complete moving targets in the two frames of images;

2. When , there may be repeated and incomplete moving targets in the two frames of images;

时，两帧图像中可能会有运动目标完全没有被拍摄到。3. When

, there may be moving objects in the two frames of images that are not captured at all.

就会避免遗漏运动目标。Obviously, in the second case, repetitive and inaccurate images of moving targets will be obtained, and in the third case, images of moving targets will be missed. According to the above analysis, it is only necessary to satisfy

It will avoid missing sports goals.

When multiple moving objects are closely placed together on the conveyor belt, the same moving object may appear repeatedly in two adjacent frames of images, thereby being repeatedly positioned, resulting in empty grabbing of parallel manipulators. In order to avoid the above situation, the present invention The embodiment provides a second decision-making condition, the second decision-making condition is specifically: every time the conveyor belt moves ds pixels, the camera shoots once, and the difference between the image coordinates x of the same moving target in two adjacent frames of images is a fixed value ds( The y value remains unchanged), using the difference method to compare the position information of the moving target obtained in the two frames of images one by one, discarding the position information of the moving target with the same y coordinate value and the difference between the x coordinate value is ds, the process is as follows :

1. Define the coordinate set of multiple moving targets obtained in the first frame image in two adjacent frames as A={( _xi , y _i )|0< _xi , y _i <M, i∈N} , the coordinate set of multiple moving targets obtained in the second frame image is B={(x _j , y _j )|0<x _j , y _j <M, j∈N};

2. Calculate (x _j -x _i ) and (y _j -y _i ), the results are divided into the following situations:

(1) When x _{j -xi} = ds and y _j -y _i = 0, it means that the jth moving object in set B is the repetition of the ith moving object in set A, and the jth moving object in set B corresponds to The location information of should be discarded;

(2) When x _{j -xi} = ds and y _j -y _i ≠ 0, it means that the jth moving target in set B is located on the same vertical line as the i-th moving target in set A, and the jth moving target in set B The location information corresponding to the moving target should be retained;

(3) When x _{j -xi} ≠ ds, it means that the jth moving object in set B is not the same moving object as the i-th moving object in set A, and the position information corresponding to the jth moving object in set B should be reserved.

Among them, M is the length of the field of view, N is the width of the field of view, the coordinates of the moving target in the image are two-dimensional coordinates, x is the abscissa of the moving target, and y is the vertical coordinate of the moving target.

105: The computer obtains the position coordinates of the moving object according to the effective position information of the moving object in step 104 and the displacement information of the conveyor belt in step 102, and saves the obtained position coordinates in the captured moving object database;

Wherein, the above-mentioned position coordinates are the uniquely determined position coordinates on the conveyor belt, and the uniquely determined position coordinates are composed of the effective position information of the moving target and the displacement information provided by the conveyor belt, specifically:

1. Merge the position information of the remaining moving objects in the coordinate set B of multiple moving objects obtained in the second frame image into the set T, T={(x _k , y _k )|0<x _k , y _k <M, k∈N};

2. Add the displacement information of the conveyor belt to the corresponding points in the set T, and update the set T to T={(x _k , y _k , c _k )|0<x _k , y _k <M, k∈N};

3. Obtain the position coordinates of the moving target during the operation of the conveyor belt.

$\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)<span\; class="notranslate">\; =</span>\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; dc</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}\end{array}\right)$

Among them, the set T is recorded as the position information set of all moving objects correctly identified, M is the length of the field of view, and N is the width of the field of view. The coordinates of the moving object in the image are two-dimensional coordinates, and x is the abscissa of the moving object , y is the ordinate of the moving target, c _k is the pulse number of the conveyor belt motor at the moment of image acquisition at this point, (x, y) ^T is the position coordinate of the moving target on the conveyor belt when the pulse number of the conveyor belt motor is c, dc is the distance traveled by the unit pulse conveyor belt.

106: The parallel manipulator sequentially extracts the position coordinates of the moving object from the grasped moving object database in step 105, and performs a grasping operation.

Specifically, the parallel manipulator sequentially extracts the position coordinates of the moving target from the database of the grasped moving target, and the position coordinates are transmitted to the control system of the parallel manipulator, and the control system instructs the parallel manipulator to grab at the position corresponding to the position coordinates Operation, after the grabbing is completed, the parallel manipulator puts the moving target into the tray on another conveyor belt.

To sum up, the embodiment of the present invention provides a method for tracking a moving object on a production line. The method collects image information of the moving object at a fixed position through a visual sensor, obtains the image coordinates of the moving object through calculation, and combines the displacement information provided by the conveyor belt. , the first decision-making condition, the second decision-making condition, transfer the position coordinates of the tracked moving target to the parallel manipulator, provide positioning information for the parallel manipulator to perform reliable and fast grabbing operations, and adapt to the requirements of the automated production line; avoid movement The target is repeatedly recognized or missed; and through the two-way feedback control between the servo motor and the visual sensor shooting frequency, the parallel manipulator meets the requirements of high-speed and real-time grasping of scattered moving targets.

The embodiment of the present invention uses a simple test to verify the effectiveness of the method provided by the embodiment of the present invention.

Taking the real object as an example, randomly select 15 positions on the conveyor belt, measure and record the distance between the relative positions (for subsequent error analysis), and place a moving target at each position, as shown in Figure 5. Analyze three consecutive frames of images, the specific parameters are set to ds=0.5M, that is, the conveyor belt takes a picture once every half of the length of the camera’s field of view, and three frames of images are obtained, see Figure 6, Figure 7, and Figure 8. 6. The conclusions in Table 1 can be obtained from the analysis of Figure 7 and Figure 8 .

See Table 1. Table 1 is the result obtained through the first decision-making condition and the second decision-making condition. After identification, a total of 15 moving targets are obtained, which is consistent with the actual situation. The relative distance between the moving targets and the distance between the moving targets recorded before the experiment The mean square error of the distance is 0.28mm. After analysis, the source of the error includes the calibration error of the camera and the slight vibration that occurs when the moving target runs synchronously on the conveyor belt, but the accuracy can be accepted by the grasping requirements of the parallel manipulator, and there is no repeated counting and loss in the whole process In the case of moving targets, the stability of the system has been well verified.

Table 1

Goal 1 Goal 2 Goal 3 Goal 4 Goal 5 Goal 6 Goal 7 x 421.868 480.747 388.846 433.041 487.104 464.979 517.54 y 22.267 29.034 64.042 72.189 73.853 116.200 122.79 pulse 30000 30000 30000 30000 30000 30000 30000

continued

Goal 8 Goal 9 Goal 10 Goal 11 Goal 12 Goal 13 Goal 14 Goal 15 399.765 481.10 434.50 503.30 483.19 443.10 501.72 433.44 123.623 42.34 63.64 67.36 102.11 44.628 74.172 91.08 30000 126000 126000 126000 126000 222000 222000 222000

Through the above-mentioned experimental verification, it can be seen that the method provided by the embodiment of the present invention is feasible, can meet the requirements of parallel manipulators for high-speed and real-time grasping of scattered moving objects, and meets the needs of practical applications.

Those skilled in the art can understand that the accompanying drawing is only a schematic diagram of a preferred embodiment, and the serial numbers of the above-mentioned embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within range.

Claims (6)

1. A method for tracking a moving target on a production line is characterized by comprising the following steps:

(1) the vision sensor collects the image of the moving target and transmits the image to the computer;

(2) the servo motor controls the running speed of the conveyor belt to obtain the displacement information of the conveyor belt;

(3) the computer receives the image in the step (1) and identifies the position information of the moving target;

(4) the computer eliminates the conditions that the moving target is omitted and repeatedly identified according to the image in the step (1), the position information in the step (3), the first decision condition and the second decision condition, and obtains effective position information of the moving target;

(5) the computer acquires the position coordinates of the moving object according to the effective position information in the step (4) and the displacement information in the step (2), and stores the position coordinates in a captured moving object database;

(6) and (5) sequentially extracting the position coordinates of the moving object from the captured moving object database in the step (5) by the parallel manipulator, and performing capturing operation.

2. The method according to claim 1, wherein the first decision condition in step (4) is specifically:

wherein ds is the number of pixels moved by the conveyor belt, M is the length of the field of view, the coordinate of the moving object in the image is a two-dimensional coordinate, x is the abscissa of the moving object, y is the ordinate of the moving object, the direction of the x coordinate is consistent with the advancing direction of the conveyor belt, the direction of the y coordinate is consistent with the width direction of the field of view,

is the maximum length of the moving object in the x-axis direction.

3. The method according to claim 1, wherein the second decision condition in step (4) is specifically:

the coordinate set of the plurality of moving objects obtained in the first frame image in the two adjacent frame images is A { (x)_i，y_i)|0＜x_i，y_iAnd (M, i belongs to N), and the coordinate set of the plurality of moving objects obtained in the second frame image is B { (x)_j，y_j)|0＜x_j，y_jLess than M, j belongs to N, and two frames of images are processed by a difference making methodComparing the position information of the moving target obtained in the image one by one;

when x is_j-x_iDs and y_j-y_iWhen the number of the motion objects is equal to 0, the j-th motion object in the set B is the repetition of the i-th motion object in the set A, and the position information corresponding to the j-th motion object in the set B is omitted;

when x is_j-x_iDs and y_j-y_iWhen the position information of the jth moving target in the set B is not equal to 0, the jth moving target in the set B and the ith moving target in the set A are positioned on the same vertical line, and the position information corresponding to the jth moving target in the set B is reserved;

when x is_j-x_iWhen the distance is not equal to ds, it is indicated that the jth moving target in the set B and the ith moving target in the set A are not the same moving target, and the position information corresponding to the jth moving target in the set B is reserved;

wherein ds is the number of pixels moved by the conveyor belt, M is the length of the field of view, N is the width of the field of view, the coordinates of the moving object in the image are two-dimensional coordinates, x is the abscissa of the moving object, and y is the ordinate of the moving object.

4. The method according to claim 1, wherein the computer in step (5) obtains the position coordinates of the moving object according to the effective position information in step (4) and the displacement information in step (2), specifically:

merging the position information of the rest moving objects in the coordinate set of the plurality of moving objects obtained in the second frame image into a set T, T { (x)_k，y_k)|0＜x_k，y_k＜M，k∈N}；

Adding the displacement information of the conveyor belt into a corresponding point in a set T, and updating the set T into T { (x)_k，y_k，c_k)|0＜x_k，y_k＜M，k∈N}；

Acquiring the position coordinates of a moving object in the running process of the conveyor belt;

$\left(\begin{array}{c}x\\ y\end{array}\right)=\left(\begin{array}{c}x=x_k+(c-c_k)\mathrm{dc}\\ y=y_k\end{array}\right)$

wherein the set T is recorded as a position information set of all correctly identified moving targets, M is the length of a view field, N is the width of the view field, the coordinates of the moving targets in the image are two-dimensional coordinates, x is the abscissa of the moving targets, y is the ordinate of the moving targets, c is the distance between the moving targets and the image, and the distance between the moving targets and the image is the distance between the moving targets_kFor this point, the number of pulses of the belt motor is transmitted at the time of image acquisition, (x, y)^TThe position coordinate of the moving object on the conveyor belt when the number of pulses of the conveyor belt motor is c, and dc is the distance of advance of the pulse conveyor belt.

5. The method according to claim 1, wherein the displacement information in step (2) is specifically:

and the displacement information of the conveyor belt at the current moment is obtained by reading the pulse numerical value in the servo motor.

6. The method of claim 1, wherein a bi-directional feedback control is provided between the servo motor and the vision sensor.

Images