Target identification and positioning device based on machine vision

Abstract

The invention discloses a target identification and positioning device based on machine vision, and relates to the technical field of identification and positioning, and the device comprises a control module which is responsible for adjusting and controlling an industrial camera; the image acquisition module is responsible for acquiring image data of a target object; the image processing module is responsible for processing the acquired image and extracting features; the identification and positioning module is responsible for identifying a target object by using the extracted feature information and determining the specific position of the target object in the image; the control module is arranged on the protection box, the exterior of the protection box is rotationally connected with a third rotating shaft, when the position of the industrial camera needs to be adjusted, the third rotating shaft is rotated to finally adjust the angle of the connecting frame for obtaining the target information, and the height of the industrial camera can be adjusted through the lifting mechanism; therefore, the visual field range of the industrial camera can be adjusted conveniently, and the method is suitable for different application scenes, so that the target object presents the optimal size and position in the image, and the positioning precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of identification and positioning, and in particular to a target identification and positioning device based on machine vision. Background Art

[0002] With the rapid development of computer vision, image processing, pattern recognition and other fields. These technologies provide a strong theoretical basis and technical support for machine vision systems, enabling machines to simulate human visual functions and identify, locate, measure and track targets in images and videos. With the continuous development of technology, target recognition and positioning devices based on machine vision have been widely used in various fields, such as industrial automation, intelligent manufacturing, intelligent logistics, unmanned driving, etc. These devices can achieve fast and accurate recognition and positioning of targets, improve production efficiency and safety, and reduce labor costs. At the same time, with the continuous development of technologies such as deep learning, the recognition accuracy and efficiency of machine vision systems will continue to improve, bringing more possibilities for future intelligent production and life.

[0003] In the prior art, it is difficult to adjust the position of the camera. Target objects of different shapes and sizes may require different shooting angles and heights to be accurately identified and located. In other words, it is difficult to ensure that the camera can capture the target object at the optimal angle and distance, and it is difficult to improve the accuracy of identification and positioning. Therefore, it is necessary to propose a target recognition and positioning device based on machine vision. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art that it is difficult to adjust the position of the camera, and target objects of different shapes and sizes may require different shooting angles and heights to be accurately identified and located, that is, it is difficult to ensure that the camera can capture the target object at the optimal angle and distance, and it is difficult to improve the accuracy of identification and positioning. A target recognition and positioning device based on machine vision is proposed to solve the shortcomings in the prior art that it is difficult to adjust the position of the camera, and target objects of different shapes and sizes may require different shooting angles and heights to be accurately identified and located. That is, it is difficult to ensure that the camera can capture the target object at the optimal angle and distance, and it is difficult to improve the accuracy of identification and positioning.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A target recognition and positioning device based on machine vision, comprising:

[0007] Control module: responsible for adjusting and controlling the industrial camera;

[0008] Image acquisition module: responsible for acquiring image data of the target object;

[0009] Image processing module: responsible for processing the collected images and extracting features;

[0010] Identification and positioning module: responsible for identifying the target object and determining its specific location in the image using the extracted feature information;

[0011] The control module is arranged on the protection box, the outside of the protection box is rotatably connected with a third rotating shaft, the outside of the third rotating shaft is fixedly connected with a third gear, the outside of the third gear is meshingly connected with an annular rack, the outside of the annular rack is fixedly connected with a rotating ring, the outside of the rotating ring is fixedly connected with a plurality of telescopic rods, the side of the telescopic rod away from the rotating ring is fixedly connected with a placing table, the outside of the placing table is fixedly installed with an industrial camera, and a lifting mechanism is arranged on the outside of the protection box, and the lifting mechanism is convenient for adjusting the height of the industrial camera;

[0012] The control module is divided into an angle adjustment unit and a height adjustment unit. The angle adjustment unit can adjust the acquisition angle of the industrial camera, and the height adjustment unit can adjust the acquisition height of the industrial camera, which can ensure that the target object is completely within the imaging range of the industrial camera, and help the industrial camera to more accurately capture the features of the target object and improve the accuracy of recognition and positioning.

[0013] The above technical solution further includes:

[0014] The image processing module uses a binary threshold processing algorithm to process the image, and its formula is:

[0015] [B(x,y)=\begin{cases}255,&\text{if}I(x,y)>T\0,&\text{if}I(x,y)\leqT\end{cases}]Where (B(x,y)) represents the pixel value of the output binary image at position ((x,y));

[0016] (I(x,y)) represents the pixel value of the input grayscale image at position ((x,y));

[0017] (T) indicates a set threshold value.

[0018] The identification and positioning module uses a support vector machine to perform identification and positioning, and the specific steps are as follows:

[0019] Data preprocessing: Preprocess the image and convert the image data into a form suitable for SVM processing;

[0020] Select kernel function: select appropriate kernel function according to data characteristics;

[0021] Training SVM model: Use training data to train the SVM model and obtain the optimal hyperplane parameters w and b by solving the optimization problem;

[0022] Model evaluation: Use a validation set or a test set to evaluate the performance of the model;

[0023] Apply the SVM model for target recognition and positioning: For new image data, first perform the same preprocessing and feature extraction, then input the extracted feature vector into the trained SVM model to obtain the category label of the sample, and after identifying the target, further determine the specific location of the target in the image.

[0024] A second motor is fixedly installed on the outside of the protection box, the output end of the second motor is fixedly connected to the third rotating shaft, the rotating ring and the protection box are rotatably connected, and a plurality of universal wheels are fixedly connected to the outside of the protection box.

[0025] The lifting mechanism includes a first motor arranged outside the protective box, a threaded rod is provided at the output end of the first motor, the external thread of the threaded rod is connected to a connecting frame, the external part of the connecting frame is fixedly connected to a second rack, the external part of the second rack is meshedly connected to a first gear, the external part of the first gear is fixedly connected to a first rotating shaft, and the first rotating shaft is rotatably connected to the protective box.

[0026] The first rotating shaft is fixedly connected to the outside with a first connecting plate, the first connecting plate is rotatably connected to the outside with a second connecting plate, the second connecting plate is rotatably connected to the outside with a detection platform, and the detection platform is rotatably connected to the placement platform.

[0027] The outside of the connecting frame is fixedly connected with a first rack, the outside of the first rack is meshingly connected with a second gear, the outside of the second gear is fixedly connected with a second rotating shaft, and the second rotating shaft is rotatably connected to the protective box.

[0028] The outside of the second gear is fixedly connected with a third connecting plate, the outside of the third connecting plate is rotatably connected with a fourth connecting plate, and the fourth connecting plate is rotatably connected to the detection platform.

[0029] The present invention has the following beneficial effects:

[0030] 1. In the present invention, when the position of the industrial camera needs to be adjusted, the third rotating axis is rotated to finally adjust the angle at which the connecting frame obtains the target information. The height of the industrial camera can be adjusted by the lifting mechanism, which is convenient for adjusting the field of view of the industrial camera and is suitable for different application scenarios, so that the target object presents the best size and position in the image, thereby improving the positioning accuracy.

[0031] 2. In the present invention, the position of the industrial camera can be adjusted through the control module, so that the image acquisition module can collect more accurate data, and then the target object can be identified and located by processing and analyzing the collected data. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a schematic diagram of the structure of a target recognition and positioning device based on machine vision proposed by the present invention;

[0033] Figure 2 It is a schematic diagram of the first structure in the present invention;

[0034] Figure 3 It is a second structural schematic diagram of the present invention;

[0035] Figure 4 for Figure 1 A schematic diagram of the structure enlargement in the middle;

[0036] Figure 5 for Figure 3 A magnified schematic diagram of the structure at B in the middle;

[0037] Figure 6 It is a system block diagram of the present invention.

[0038] In the figure: 1. protective box; 2. universal wheel; 3. testing table; 4. first connecting plate; 5. second connecting plate; 6. first rotating shaft; 7. first gear; 8. first rack; 9. third connecting plate; 10. fourth connecting plate; 11. second gear; 12. second rotating shaft; 13. telescopic rod; 14. rotating ring; 15. connecting frame; 16. first motor; 17. second rack; 18. threaded rod; 19. placing table; 20. industrial camera; 21. second motor; 22. third rotating shaft; 23. third gear; 24. annular rack. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] like Figure 1-6 As shown, a target recognition and positioning device based on machine vision includes:

[0041] Control module: responsible for adjusting and controlling the industrial camera 20;

[0042] Image acquisition module: responsible for acquiring image data of the target object;

[0043] Image processing module: responsible for processing the collected images and extracting features;

[0044] Identification and positioning module: responsible for identifying the target object and determining its specific location in the image using the extracted feature information;

[0045] The control module is arranged on the protection box 1, the outside of the protection box 1 is rotatably connected with a third rotating shaft 22, the outside of the third rotating shaft 22 is fixedly connected with a third gear 23, the outside of the third gear 23 is meshingly connected with an annular rack 24, the outside of the annular rack 24 is fixedly connected with a rotating ring 14, the outside of the rotating ring 14 is fixedly connected with a plurality of telescopic rods 13, the side of the telescopic rods 13 away from the rotating ring 14 is fixedly connected with a placement table 19, the outside of the placement table 19 is fixedly installed with an industrial camera 20, and a lifting mechanism is arranged on the outside of the protection box 1, and the lifting mechanism is convenient for adjusting the height of the industrial camera 20;

[0046] The control module is divided into an angle adjustment unit and a height adjustment unit. The angle adjustment unit can adjust the acquisition angle of the industrial camera 20, and the height adjustment unit can adjust the acquisition height of the industrial camera 20. This can ensure that the target object is completely within the imaging range of the industrial camera 20, which helps the industrial camera 20 to more accurately capture the features of the target object and improve the accuracy of recognition and positioning.

[0047] The working principle of a target recognition and positioning device based on machine vision proposed by the present invention is that the control module can adjust the industrial camera 20 to help the industrial camera 20 better collect image data of the target object, the image acquisition module can obtain the image or video data of the target object, and the data processing module will pre-process the collected image, such as denoising, enhancement, filtering and other operations, and extract feature information related to the target object from the processed image to improve the accuracy and efficiency of subsequent recognition and positioning;

[0048] When the angle of the industrial camera 20 needs to be adjusted, the third rotating shaft 22 is rotated, and the third rotating shaft 22 will drive the third gear 23 fixedly connected to the outside to rotate, and the third gear 23 will cause the annular rack 24 meshing with the outside to rotate, and when the annular rack 24 rotates, the rotating ring 14 fixedly connected to the outside will rotate outside the protective box 1, and a plurality of telescopic rods 13 are fixedly connected to the outside of the rotating ring 14, and the outside of the telescopic rods 13 are fixedly connected to the placement table 19, that is, when the rotating ring 14 rotates, it will drive the telescopic rods 13 to rotate, and the telescopic rods 13 will drive the placement table 19 to rotate, that is, the angle of the industrial camera 20 fixedly installed on the outside of the placement table 19 to obtain the image data information of the target object can be changed, and the lifting mechanism is convenient for adjusting the angle of the industrial camera 20, and can ensure that the target object is completely located within the imaging range of the industrial camera 20.

[0049] In one embodiment, for the image processing module, the image processing module applies a binary threshold processing algorithm to process the image, and the formula is:

[0050] [B(x,y)=\begin{cases}255,&\text{if}I(x,y)>T\0,&\text{if}I(x,y)\leqT\end{cases}]Where (B(x,y)) represents the pixel value of the output binary image at position ((x,y));

[0051] (I(x,y)) represents the pixel value of the input grayscale image at position ((x,y));

[0052] (T) indicates a set threshold value.

[0053] In one embodiment, for the recognition and positioning module, the recognition and positioning module uses a support vector machine to perform recognition and positioning, and the specific steps are:

[0054] Data preprocessing: Preprocess the image and convert the image data into a form suitable for SVM processing;

[0055] Select kernel function: select appropriate kernel function according to data characteristics;

[0056] Training SVM model: Use training data to train the SVM model and obtain the optimal hyperplane parameters w and b by solving the optimization problem;

[0057] Model evaluation: Use a validation set or a test set to evaluate the performance of the model;

[0058] Apply the SVM model for target recognition and positioning: For new image data, first perform the same preprocessing and feature extraction, then input the extracted feature vector into the trained SVM model to obtain the category label of the sample, and after identifying the target, further determine the specific location of the target in the image.

[0059] In one embodiment, for the protective box 1, a second motor 21 is fixedly installed on the outside of the protective box 1, the output end of the second motor 21 is fixedly connected to the third rotating shaft 22, the rotating ring 14 and the protective box 1 are rotatably connected, and a plurality of universal wheels 2 are fixedly connected to the outside of the protective box 1.

[0060] In one embodiment, for the above-mentioned lifting mechanism, the lifting mechanism includes a first motor 16 arranged outside the protective box 1, and the output end of the first motor 16 is provided with a threaded rod 18, the external thread of the threaded rod 18 is connected to the connecting frame 15, the external part of the connecting frame 15 is fixedly connected to the second rack 17, the external part of the second rack 17 is meshedly connected to the first gear 7, the external part of the first gear 7 is fixedly connected to the first rotating shaft 6, the first rotating shaft 6 and the protective box 1 are rotatably connected, the external part of the first rotating shaft 6 is fixedly connected to the first connecting plate 4, the external part of the first connecting plate 4 is rotatably connected to the second connecting plate 5, the external part of the second connecting plate 5 is rotatably connected to the detection table 3, and the detection table 3 and the placement table 19 are rotatably connected.

[0061] The working principle of a target recognition and positioning device based on machine vision proposed by the present invention is to start the first motor 16 fixedly installed on the outside of the protective box 1, and the first motor 16 will cause the threaded rod 18 set at the output end to rotate. When the threaded rod 18 rotates, the connecting frame 15 with an external thread connection will move. When the connecting frame 15 moves, it will drive the second rack 17 fixedly connected to the outside to move. When the second rack 17 moves, it will drive the first gear 7 engaged with the outside to rotate. When the first gear 7 rotates, the first rotating shaft 6 fixedly connected to the outside will rotate. When the first rotating shaft 6 rotates, it will drive the first connecting plate 4 fixedly connected to the outside to rotate. When the first connecting plate 4 rotates, it will cause the second connecting plate 5 with an external rotation connection to rotate. The second connecting plate 5 is rotatably connected to the detection platform 3, that is, when the second connecting plate 5 rotates, the height of the detection platform 3 can be changed, that is, the height of the industrial camera 20 can be changed.

[0062] In one embodiment, for the above-mentioned connecting frame 15, the outside of the connecting frame 15 is fixedly connected to the first rack 8, the outside of the first rack 8 is meshingly connected to the second gear 11, the outside of the second gear 11 is fixedly connected to the second rotating shaft 12, and the second rotating shaft 12 is rotatably connected to the protective box 1.

[0063] In one embodiment, for the second gear 11 , the outside of the second gear 11 is fixedly connected to the third connecting plate 9 , the outside of the third connecting plate 9 is rotatably connected to the fourth connecting plate 10 , and the fourth connecting plate 10 is rotatably connected to the detection platform 3 .

[0064] The working principle of the target recognition and positioning device based on machine vision proposed by the present invention is that when the connecting frame 15 moves, the first rack 8 will be driven to rotate at the same time. The first rack 8 and the second gear 11 are meshingly connected, and the first rack 8 and the second rack 17 are in opposite directions, that is, the first gear 7 and the second gear 11 will rotate in opposite directions, that is, the second gear 11 will drive the second rotating shaft 12 fixedly connected to the outside to rotate, the second rotating shaft 12 will rotate the first rack 8 connected to the outside to rotate, and the first rack 8 will rotate the third connecting plate 9 connected to the outside to rotate, so as to keep the detection platform 3 stable when it is raised or lowered.

[0065] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A target recognition and positioning device based on machine vision, characterized in that: include: Control module: responsible for regulating and controlling the industrial camera (20); Image acquisition module: responsible for acquiring image data of the target object; Image processing module: responsible for processing the collected images and extracting features; Identification and positioning module: responsible for identifying the target object and determining its specific location in the image using the extracted feature information; The control module is arranged on the protection box (1), the outside of the protection box (1) is rotatably connected to a third rotating shaft (22), the outside of the third rotating shaft (22) is fixedly connected to a third gear (23), the outside of the third gear (23) is meshingly connected to an annular rack (24), the outside of the annular rack (24) is fixedly connected to a rotating ring (14), the outside of the rotating ring (14) is fixedly connected to a plurality of telescopic rods (13), the side of the telescopic rods (13) away from the rotating ring (14) is fixedly connected to a placement table (19), the outside of the placement table (19) is fixedly installed with an industrial camera (20), and the outside of the protection box (1) is provided with a lifting mechanism, which facilitates the adjustment of the height of the industrial camera (20); The control module is divided into an angle adjustment unit and a height adjustment unit. The angle adjustment unit can adjust the acquisition angle of the industrial camera (20), and the height adjustment unit can adjust the acquisition height of the industrial camera (20), thereby ensuring that the target object is completely within the imaging range of the industrial camera (20), helping the industrial camera (20) to more accurately capture the features of the target object and improve the accuracy of recognition and positioning.

2. The target recognition and positioning device based on machine vision according to claim 1, characterized in that: The image processing module uses a binary threshold processing algorithm to process the image, and its formula is: [B(x,y)=\begin{cases}255,&\text{if}I(x,y)>T\0,&\text{if}I(x,y)\leq T\end{cases}]Where (B(x,y)) represents the pixel value of the output binary image at position ((x,y)); (I(x,y)) represents the pixel value of the input grayscale image at position ((x,y)); (T) indicates a set threshold value.

3. The target recognition and positioning device based on machine vision according to claim 1, characterized in that: The identification and positioning module uses a support vector machine to perform identification and positioning, and the specific steps are as follows: Data preprocessing: Preprocess the image and convert the image data into a form suitable for SVM processing; Select kernel function: select appropriate kernel function according to data characteristics; Training SVM model: Use training data to train the SVM model and obtain the optimal hyperplane parameters w and b by solving the optimization problem; Model evaluation: Use a validation set or a test set to evaluate the performance of the model; Apply the SVM model for target recognition and positioning: For new image data, first perform the same preprocessing and feature extraction, then input the extracted feature vector into the trained SVM model to obtain the category label of the sample, and after identifying the target, further determine the specific location of the target in the image.

4. The target recognition and positioning device based on machine vision according to claim 1, characterized in that: A second motor (21) is fixedly mounted on the outside of the protection box (1); an output end of the second motor (21) is fixedly connected to a third rotating shaft (22); the rotating ring (14) is rotatably connected to the protection box (1); and a plurality of universal wheels (2) are fixedly connected to the outside of the protection box (1).

5. The target recognition and positioning device based on machine vision according to claim 1, characterized in that: The lifting mechanism comprises a first motor (16) arranged outside the protective box (1), a threaded rod (18) is provided at the output end of the first motor (16), the external thread of the threaded rod (18) is connected to a connecting frame (15), the external part of the connecting frame (15) is fixedly connected to a second rack (17), the external part of the second rack (17) is meshingly connected to a first gear (7), the external part of the first gear (7) is fixedly connected to a first rotating shaft (6), and the first rotating shaft (6) is rotatably connected to the protective box (1).

6. The target recognition and positioning device based on machine vision according to claim 5, characterized in that: The first rotating shaft (6) is fixedly connected to the outside of a first connecting plate (4), the first connecting plate (4) is rotatably connected to the outside of a second connecting plate (5), the second connecting plate (5) is rotatably connected to the outside of a detection platform (3), and the detection platform (3) is rotatably connected to the placement platform (19).

7. The target recognition and positioning device based on machine vision according to claim 5, characterized in that: The outside of the connecting frame (15) is fixedly connected to a first rack (8), the outside of the first rack (8) is meshingly connected to a second gear (11), the outside of the second gear (11) is fixedly connected to a second rotating shaft (12), and the second rotating shaft (12) is rotatably connected to the protective box (1).

8. The target recognition and positioning device based on machine vision according to claim 7, characterized in that: The outside of the second gear (11) is fixedly connected to a third connecting plate (9), the outside of the third connecting plate (9) is rotatably connected to a fourth connecting plate (10), and the fourth connecting plate (10) is rotatably connected to the detection platform (3).

Images