A strabismus camera vision recognition system

Abstract

The application relates to a strabismus camera visual identification system in the technical field of welding, which comprises a welding device, a camera is arranged on one side of the top of the welding device, a clamping device is arranged on the bottom of the welding device in the extension direction of a vertical central axis, and a lifting platform is slidably connected in the clamping device; the application further comprises a PLC controller, the PLC controller comprises a lifting module for controlling the vertical sliding of the lifting platform in the clamping device, and comprises an origin identification module for identifying the origin position of the top surface of a single-section fan blade, the lifting platform is electrically connected with the lifting module, and the camera is electrically connected with the origin identification module. The application can improve the efficiency and accuracy of welding the middle section of the fan blade.

Description

Technical Field

[0001] This application relates to the field of welding technology, and in particular to a squinting camera visual recognition system. Background Technology

[0002] Cross-flow fan impellers, as a common ventilation equipment component, are widely used in household appliances such as air conditioners and fans. Due to their unique multi-blade design and elongated cylindrical structure, they effectively improve airflow efficiency and stability. In recent years, with the continuous growth of the household appliance market and technological advancements, higher demands have been placed on the manufacturing process of cross-flow fan impellers. Especially in the welding process, achieving efficient and precise origin positioning has become one of the key technical challenges.

[0003] In existing technologies, cameras are typically used for identification to ensure the precise assembly of individual blade sections in a cross-flow wind turbine. For example, a robotic arm moves the camera to the top of the blade section to take a picture and identify its position before the camera is moved out of the welding station for welding. While this method can meet the requirement of origin point identification to some extent, it has many inconveniences in practical applications.

[0004] However, in practice, the above method involves frequent camera movement, which not only increases the complexity and cost of the system but also significantly reduces the overall efficiency of welding production. Therefore, it is necessary to improve the efficiency and accuracy of origin point identification during the welding process of cross-flow wind turbines. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this application provides a slanted camera visual recognition system that can improve the efficiency and accuracy of welding the middle section of a fan blade.

[0006] This application provides a visual recognition system using a squinting camera, employing the following technical solution:

[0007] A slant-view camera visual recognition system, comprising:

[0008] A welding device, wherein a camera is provided on one side of the top of the welding device, and a clamping device is provided at the bottom of the welding device in the direction of the extension of the vertical central axis, and a lifting platform is slidably connected inside the clamping device;

[0009] The PLC controller includes a lifting module for controlling the vertical sliding of the lifting platform within the clamping device, and an origin recognition module for identifying the origin position of the top surface of a single fan blade. The lifting platform is electrically connected to the lifting module, and the camera is electrically connected to the origin recognition module.

[0010] By adopting the above technical solution, the lifting device accommodates a single section of the fan blade inside the clamping device, shortening the height of the multi-section fan blade and leaving more space for the camera to capture the top surface of the fan blade. The camera of the welding device is angled, reducing the process of the camera moving to the top of the welding device and then moving out, shortening the time for the camera to confirm the position of the middle section of the fan blade. Furthermore, the origin position of the top surface of the single section of the fan blade is identified, and the angle of the single section of the fan blade is confirmed by the origin position, providing accurate data for the subsequent adjustment of the angle of the single section of the fan blade, thereby improving the efficiency and accuracy of welding single section fan blades.

[0011] This application further provides that the PLC controller also includes an origin correction module for correcting the origin shape of the top surface of the fan blade located in the clamping device, and the origin correction module is electrically connected to the origin identification module.

[0012] By adopting the above technical solution, the origin correction module can correct the shape of the origin that is deformed due to the camera shooting the top of the middle section of the wind turbine blade at an angle. The corrected origin position reflects the angle deviation of the middle section of the wind turbine blade more accurately.

[0013] This application further includes a wind turbine blade database, which stores wind turbine blade data of different diameters and heights captured by a camera. The PLC controller also includes a component segmentation module for cutting the side and top surfaces of the wind turbine blades. The component segmentation module is electrically connected to the origin recognition module, the origin correction module, and the wind turbine blade database.

[0014] By adopting the above technical solution, while being able to adapt to the size and height of wind blades of various dimensions, the component segmentation module can separate the top surface and side surface of the wind blade after the image is captured. The origin recognition module and the origin correction module analyze the origin of the top surface of the wind blade after segmentation, and narrowing the range of origin recognition can improve the accuracy of origin recognition and improve the accuracy of welding the middle section of the wind blade.

[0015] This application further provides that the PLC controller also includes a coordinate acquisition module for generating and saving coordinate data of the top surface of the fan blade, and the coordinate acquisition module is electrically connected to the component segmentation module and the origin correction module.

[0016] By adopting the above technical solution and using coordinate data instead of image data, the space required to store different types of wind turbine blade information can be reduced, the data reading speed can be improved, and the efficiency of identifying and welding wind turbine blade sections can be increased.

[0017] This application further provides that: the PLC controller also includes a height acquisition module for obtaining the height of the wind turbine blade based on the coordinate data of the top surface of the blade, and the height acquisition module is electrically connected to the coordinate acquisition module, the lifting module, and the camera.

[0018] By adopting the above technical solution, when welding the top surfaces of the combined wind blades, if the height of each wind blade section is the same, but the height of the lifting platform descends differently each time, and / or the height of the lifting platform descends the same each time, but the heights of adjacent welded wind blades are different, the edge coordinates of the top surface of the wind blades will deviate. Based on the entered coordinate data of the top surface of the wind blades, the actual height difference between the current wind blade and the previous wind blade is obtained. Then, the height of the lifting platform is adjusted by the lifting module to make the coordinate data of the top surface of the wind blade of the current welding station consistent with the top surface of the wind blade being welded, thereby improving the recognition accuracy of the origin recognition module.

[0019] This application further specifies that the shooting direction of the camera is at an angle of 30°-60° to the extension direction of the vertical central axis of the clamping device.

[0020] By adopting the above technical solution, the 30°-60° angle setting allows the camera to effectively avoid obstruction by the welding device without interfering with the welding operation, thereby achieving efficient and accurate origin recognition and improving welding efficiency and accuracy.

[0021] In summary, this application has the following beneficial effects:

[0022] This application uses an oblique-view camera to identify the origin of the top surface of the middle section, transforming the identified elliptical cross-section into a circular cross-section. This significantly improves the speed and accuracy of origin identification, thereby increasing welding production efficiency. At the same time, when the height of the fan blade or lifting platform changes, it can quickly and accurately correct the origin position, ensuring the stability and accuracy of the welding process. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the module connections in this application.

[0024] Figure 2 This is a schematic diagram of the overall structure of this application.

[0025] Reference numerals: 1. Welding device; 2. Camera; 3. Clamping device; 30. Lifting platform; 4. Origin recognition module; 5. Lifting module; 6. Origin correction module; 7. Component segmentation module; 8. Coordinate acquisition module; 9. Height acquisition module. Detailed Implementation

[0026] The following is in conjunction with the appendix Figures 1 to 2 This application will be described in further detail.

[0027] refer to Figure 1 and Figure 2In this embodiment, a squinting camera visual recognition system includes a welding device 1, a camera 2 fixedly connected to one side of the top of the welding device 1, and a clamping device 3 provided at the bottom of the welding device 1 in the direction of the extension of the vertical central axis. The clamping device 3 is used to clamp a lifting platform 30 that is slidably connected inside the clamping device 3. The squinting camera visual recognition system also includes a PLC controller, which includes a lifting module 5. The lifting module 5 is used to control the lifting platform 30 to slide vertically within the clamping device 3. Meanwhile, since each blade of the wind turbine has a different angle, and the angle will shift after welding the wind turbine assembly, the angle of the wind turbine to be inserted needs to be corrected when plugging into the insertion hole on the top of the wind turbine assembly. The PLC controller includes an origin identification module 4, which is used to identify the origin position of the top surface of a single wind turbine section. The lifting platform 30 is electrically connected to the lifting module 5, so that the lifting platform 30 can sink the wind turbine assembly, which is vertical after welding, into the clamping device 3, thereby reducing the height of the welding station in the welding device 1, reducing the part of the side of the wind turbine that occupies the camera 2, leaving more shooting range of the camera 2 on the top surface of the wind turbine, so that the origin identification module 4, which is electrically connected to the camera 2, can more accurately identify the origin used to locate the wind turbine angle. The camera 2 of the welding device 1 adopts an oblique shooting method, which reduces the process of moving the camera 2 to the top of the welding device 1 and then moving it out, shortening the time for the camera 2 to confirm the position of the middle section of the blade. Then, through the origin recognition module 4, the origin position of the top surface of the single blade is further identified. By the relative position of the origin of the single blade to be welded and the origin of the blade assembly, the angle of the single blade that needs to be corrected is confirmed, providing accurate data for the subsequent adjustment of the angle of the single blade to be welded, thereby improving the efficiency and accuracy of welding the single blade.

[0028] Furthermore, the PLC controller also includes an origin correction module 6, which is used to correct the shape of the origin point on the top surface of the fan blade located in the clamping device 3. The origin correction module 6 is electrically connected to the origin recognition module 4. In this embodiment, the top surface of the fan blade is circular, and the origin point of the top surface is also circular. When the top surface of the fan blade is captured by the camera 2 at an angle, both the top surface of the fan blade and the origin point appear elliptical in the acquired image. Therefore, the origin correction module 6 is needed to correct the origin point of the fan blade to obtain a more accurate origin point position. The corrected origin point position reflects the angle deviation of the middle section of the fan blade more accurately. In one embodiment, the origin correction module 6 preprocesses the image of the origin captured at an angle, including grayscale conversion, noise reduction, and edge detection, to extract a clear elliptical outline. Then, it uses ellipse fitting algorithms such as least squares to extract the geometric parameters of the ellipse from the edge information, including the center point, major axis, minor axis, and rotation angle. Next, it performs rotation correction on the image based on the ellipse's rotation angle, aligning its principal axis with the horizontal or vertical direction. Finally, it performs non-uniform scaling on the image according to the ratio of the major and minor axes, adjusting the ellipse to a circle. The excess portion is then cropped, and the corrected circular image is output, representing the shape after origin correction. Similarly, the elliptical top surface of the wind turbine blade is also corrected to a circle, and the position of the origin is marked with the center of the corrected top surface of the wind turbine blade as the origin of the coordinate system.

[0029] Furthermore, it also includes a wind turbine blade database, which stores wind turbine blade data of different diameters and heights captured by camera 2. The PLC controller also includes a component segmentation module 7, which is used to cut the side and top surfaces of the wind turbine blades. While adapting to wind turbine blades of various sizes and heights and improving the flexibility and operability of the welding device 1, the component segmentation module 7 is electrically connected to the origin recognition module 4, the origin correction module 6, and the wind turbine blade database. The component segmentation module 7 can separate the captured top and side surfaces of the wind turbine blades, so that the origin recognition model only identifies the top surface of the wind turbine blades.

[0030] Furthermore, the PLC controller also includes a coordinate acquisition module 8, which generates and saves the coordinate data of the top surface of the wind turbine blade. The coordinate data is stored in the wind turbine blade database. The coordinate acquisition module 8 is electrically connected to the component segmentation module 7 and the origin correction module 6. In this embodiment, by locating the major axis of the elliptical top surface of the wind turbine blade, taking the midpoint of the major axis as the coordinate origin, and then confirming the coordinates of the entire elliptical surface based on the coordinates of the upper half of the major axis, the wind turbine blade data is confirmed. By using the coordinate data of the wind turbine blade instead of its image data, the space required to store different types of wind turbine blade information in the database can be reduced, and the data reading speed can be improved, thereby increasing the efficiency of identifying and welding the middle section of the wind turbine blade.

[0031] Furthermore, the PLC controller also includes a height acquisition module 9, which is used to acquire the height of the fan blade based on the coordinate data of the top surface of the fan blade. The height acquisition module 9 is electrically connected to the coordinate acquisition module 8, the lifting module 5, and the camera 2. When welding assembled fan blades with identical top surfaces, there are situations where each fan blade section has the same height, but the lifting platform 30 descends at different heights each time, and / or the lifting platform 30 descends at the same height each time, but adjacent fan blades being welded have different heights. These situations will cause significant changes in the origin's position during the welding of a single fan blade section because the edge coordinate data of the fan blade's top surface acquired by the coordinate acquisition module 8 are not identical. To reduce the change in the origin's position and minimize the angle adjustment required for the single fan blade section to be welded, the radius of the fan blade's top surface is input before welding, and the standard coordinate data of the deformed ellipse in the fan blade database is matched based on the radius data. When the elliptical surface of the top surface of the current wind turbine assembly captured by camera 2 is inconsistent with the standard coordinate data, the standard height of the top of the standard wind turbine assembly protruding from the welding station is obtained to adjust the height of the lifting platform 30. The camera 2 is used for continuous monitoring so that the coordinates of the ellipse formed by the top surface of the current wind turbine assembly protruding from the welding station in camera 2 are consistent with the standard coordinate data, thereby improving the recognition accuracy of the origin recognition module 4.

[0032] Furthermore, the shooting direction of the camera 2 is set at an angle of 30°-60° to the extension direction of the vertical central axis of the clamping device 3. This allows for effective avoidance of obstruction by the welding device 1 without interfering with the welding operation, achieving efficient and accurate origin identification and improving welding efficiency and accuracy. Preferably, in this embodiment, the shooting direction of the camera 2 is set at 45° to the extension direction of the vertical central axis of the clamping device 3.

[0033] The implementation principle of this application embodiment is as follows: After the camera 2 acquires the image of the wind turbine assembly, the side and top surfaces of the single wind turbine protruding from the welding station are divided by the component segmentation module 7. Then, the coordinate data of the top surface of the wind turbine before welding is acquired by the coordinate acquisition module 8. The acquired coordinate data is compared with the standard coordinate data stored in the wind turbine database. When they are inconsistent, the height of the lifting platform 30 is adjusted according to the standard height of the top of the standard wind turbine assembly protruding from the welding station to correct the protrusion height to the standard height. Then, the origin is identified by the origin recognition module 4.

[0034] When the origin identification module 4 obtains the origin on the top surface of the wind turbine assembly, the origin correction module 6 corrects the origin to a circle, and at the same time corrects the elliptical top surface of the wind turbine to a circle. Using the center of the circle on the corrected top surface of the wind turbine as the origin of the coordinate system, the position of the origin is marked, providing accurate coordinate data to adjust the angle of the origin of the subsequent single wind turbine section to be welded.

[0035] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A visual recognition system using a squinting camera, characterized in that, include: A welding device (1) is provided with a camera (2) on one side of the top of the welding device (1), and a clamping device (3) is provided at the bottom of the welding device (1) in the direction of the extension of the vertical central axis. A lifting platform (30) is slidably connected inside the clamping device (3). The PLC controller includes a lifting module (5) for controlling the vertical sliding of the lifting platform (30) within the clamping device (3), and an origin recognition module (4) for identifying the origin position of the top surface of a single wind blade. The lifting platform (30) is electrically connected to the lifting module (5), and the camera (2) is electrically connected to the origin recognition module (4).

2. The visual recognition system for a squinting camera according to claim 1, characterized in that, The PLC controller also includes an origin correction module (6) for correcting the origin shape of the top surface of the fan blade located in the clamping device (3), and the origin correction module (6) is electrically connected to the origin identification module (4).

3. The visual recognition system for a squinting camera according to claim 2, characterized in that, It also includes a wind turbine database, which stores wind turbine data of different diameters and heights captured by the camera (2). The PLC controller also includes a component segmentation module (7) for cutting the side and top surfaces of the wind turbine blades. The component segmentation module (7) is electrically connected to the origin recognition module (4), the origin correction module (6), and the wind turbine database.

4. The squinting camera visual recognition system according to claim 3, characterized in that, The PLC controller also includes a coordinate acquisition module (8) for generating and saving coordinate data of the top surface of the wind turbine blades. The coordinate acquisition module (8) is electrically connected to the component segmentation module (7) and the origin correction module (6).

5. A visual recognition system for a squinting camera according to claim 4, characterized in that, The PLC controller also includes a height acquisition module (9) for obtaining the height of the wind turbine blade based on the coordinate data of the top surface of the blade. The height acquisition module (9) is electrically connected to the coordinate acquisition module (8), the lifting module (5), and the camera (2).

6. The visual recognition system for a squinting camera according to claim 1, characterized in that, The shooting direction of the camera (2) is set at an angle of 30°-60° to the extension direction of the vertical central axis of the clamping device (3).

Images