CN-121982049-A - Automatic centering control method and system for servo rotary table plate based on vision

CN121982049ACN 121982049 ACN121982049 ACN 121982049ACN-121982049-A

Abstract

The application provides a vision-based automatic centering control method and a vision-based automatic centering control system for a servo rotary table plate material, and relates to the technical field of stamping automation, wherein the method comprises the steps of obtaining visible light and near infrared image data of the plate material when the plate material enters a vision field; the method comprises the steps of carrying out pixel-level fusion on two image data to extract contour point cloud information and generate a digital image stream, carrying out edge positioning on the digital image stream to obtain initial pose data of a plate material and further generate target pose parameters, comparing the target pose parameters with standard pose parameters to obtain a plurality of deviation values, carrying out motion decoupling on the deviation values to generate position compensation instructions corresponding to units, finally sending the position compensation instructions to a servo driver, and controlling servo motors corresponding to the units to execute position compensation actions by the servo driver to enable actual pose of the plate material to be consistent with the standard pose parameters, and finishing centering control. The application improves the adaptability and the cooperative precision of the centering control of the plate.

Inventors

LI ZHAOXU
ZHAN QINGSHAN
TAO WEI
Liang Jinshuai
ZHAO XIPENG

Assignees

济南昊中自动化有限公司

Dates

Publication Date: 20260505
Application Date: 20260401

Claims (10)

1. A vision-based automatic centering control method for a servo rotary table plate is characterized by comprising the following steps: When a sheet material enters a preset visual field, acquiring first image data of the sheet material in a visible light wave band and second image data of the sheet material in a near infrared wave band; performing pixel-level fusion on the first image data and the second image data to generate synthetic image data, and extracting contour point cloud information from the synthetic image data; Generating a digital image stream in the process of moving the plate along the conveying direction according to the definition evaluation result of the contour point cloud information; Performing edge positioning on the digital image stream by adopting a sub-pixel edge extraction algorithm to obtain initial pose data of the plate material, and generating target pose parameters by utilizing a camera distortion model based on the initial pose data; Comparing the target pose parameter with a preset standard pose parameter based on a multi-target cooperative servo control technology to obtain a plurality of deviation values, and performing motion decoupling on the deviation values to respectively generate position compensation instructions corresponding to an integral rotation unit, a translation unit and an independent rotation unit, wherein the deviation values comprise an X-axis deviation value, a Y-axis deviation value and a rotation angle deviation value; And the position compensation instruction is issued to a servo driver, and the servo driver controls the servo motors corresponding to the integral rotating unit, the translation unit and the independent rotating unit to execute position compensation according to the cooperative motion path, so that the actual pose of the plate material is consistent with the standard pose parameter, and centering control is completed.
2. The method of claim 1, wherein the performing edge location on the digital image stream using a subpixel edge extraction algorithm to obtain initial pose data of the slab, and generating target pose parameters using a camera distortion model based on the initial pose data, comprises: Identifying a pixel set with abrupt change of gray values in an image for each frame of image in the digital image stream, and determining the area where the pixel set is located as an edge area; in the edge area, determining the pixel level edge position of the edge according to the distribution trend of the gray values, and carrying out interpolation operation according to the gray values of pixels around the pixel level edge position to obtain the sub-pixel level coordinates of the edge in an image coordinate system; Fitting out contour edges of the plate in an image coordinate system according to sub-pixel level coordinates of all edges; Calculating geometrical center coordinates of the plate in an image coordinate system and deflection angles of a main shaft of the contour edge relative to image coordinate axes according to the contour edge, and combining the geometrical center coordinates and the deflection angles to generate initial pose data; And compensating the initial pose data through a plurality of parameters of the camera distortion model to generate target pose parameters.
3. The method of claim 2, wherein compensating the image coordinates in the initial pose data by the plurality of parameters of the camera distortion model to generate target pose parameters comprises: taking the geometric center coordinates in the initial pose data as coordinates to be compensated, and taking the deflection angles in the initial pose data as angles to be compensated; establishing a mapping relation from an image coordinate system to an ideal coordinate system according to radial distortion parameters and tangential distortion parameters in the camera distortion model; correcting the coordinate to be compensated based on the mapping relation to obtain a corrected intermediate coordinate; According to external conversion parameters in the camera distortion model, establishing a conversion relation from an ideal coordinate system to a real physical coordinate system, wherein the external conversion parameters comprise a rotation matrix and a translation vector; Substituting the corrected intermediate coordinates into the conversion relation, and performing rigid transformation on the corrected intermediate coordinates through the rotation matrix and the translation vector to obtain compensated geometric center coordinates; substituting the angle to be compensated into the conversion relation, and carrying out rotary transformation on the angle to be compensated through the rotary matrix to obtain a compensated deflection angle; And generating target pose parameters based on the compensated geometric center coordinates and the compensated deflection angles.
4. The method according to claim 1, wherein the comparing the target pose parameter with a preset standard pose parameter based on the multi-target cooperative servo control technology to obtain a plurality of deviation values, and performing motion decoupling on the plurality of deviation values to generate position compensation instructions corresponding to the integral rotation unit, the translation unit and the independent rotation unit respectively, includes: Respectively carrying out difference value calculation on the X coordinate, the Y coordinate and the rotation angle in the target pose parameter and the standard X coordinate, the standard Y coordinate and the standard rotation angle in the standard pose parameter to obtain an X-axis deviation value, a Y-axis deviation value and a rotation angle deviation value; Obtaining servo control parameters, wherein the servo control parameters comprise coupling relation parameters and motion limiting parameters; According to the coupling relation parameters, resolving the X-axis deviation value, the Y-axis deviation value and the rotation angle deviation value to obtain initial to-be-compensated quantities corresponding to the units, inputting the initial to-be-compensated quantities corresponding to the units into a pre-constructed fuzzy reasoning rule base for fuzzy reasoning to obtain corrected to-be-compensated quantities corresponding to the units, wherein a fuzzy mapping relation between the to-be-compensated quantities and correction coefficients is stored in the fuzzy reasoning rule base; and based on the motion amplitude limiting parameters, respectively carrying out amplitude limiting treatment on the corrected to-be-compensated quantity corresponding to each unit to obtain the compensation quantity corresponding to each unit, and respectively converting the compensation quantity corresponding to each unit into a corresponding position compensation instruction according to a preset servo control period.
5. The method according to claim 1, wherein generating a digital image stream during the movement of the sheet in the conveying direction according to the definition evaluation result of the contour point cloud information comprises: Acquiring the gray scale change rate of each edge point in the contour point cloud information, and counting the gray scale change rates of all the edge points to obtain an average value of the gray scale change rates; Comparing the average value of the gray level change rate with a preset change rate threshold value to obtain a definition evaluation result; When the definition evaluation result indicates that the average value of the gray level change rate is smaller than the change rate threshold, adjusting the focal length parameter of the camera optical system and the light source angle of the illumination system until the average value of the gray level change rate corresponding to the re-extracted contour point cloud information reaches the change rate threshold; Taking the focal length parameter and the light source angle after the adjustment as fixed parameters, triggering a camera to continuously collect multi-frame plate images according to a set time interval in the process of moving the plate along the conveying direction; Inputting the multi-frame plate images into a self-organizing feature mapping network for image screening to obtain a plurality of image frames, and arranging the image frames according to the sequence of acquisition time to form a digital image stream.
6. The method of claim 1, wherein the pixel-wise fusing the first image data with the second image data to generate composite image data and extracting contour point cloud information from the composite image data comprises: For each pixel position in the first image data and the second image data, respectively calculating the discrete degree of gray values in a local area with the pixel as a center to obtain a first discrete degree corresponding to each pixel position in the first image data and a second discrete degree corresponding to each pixel position in the second image data; comparing the first discrete degree with the second discrete degree aiming at the same pixel position, and selecting the gray value of the pixel position in the image data with larger discrete degree as the gray value of the corresponding pixel position in the synthesized image data; traversing all pixel positions to obtain composite image data formed by gray values of all pixel positions, wherein each pixel is used as a target pixel in the composite image data, and the difference value between the gray value of the target pixel and the gray value of each adjacent pixel is calculated; and jointly using adjacent pixels with the difference value larger than a preset difference value threshold value and target pixels corresponding to the adjacent pixels as edge pixels, and combining the coordinate positions of all the edge pixels in the synthesized image data to obtain contour point cloud information.
7. The method according to claim 1, wherein after the position compensation command is issued to the servo driver, the servo driver controls the servo motors corresponding to the integral rotating unit, the translation unit and the independent rotating unit to execute the position compensation according to the cooperative motion path, so that the actual pose of the plate material is consistent with the standard pose parameter, and after the centering control is completed, the method further comprises: acquiring the stay position coordinates of the plate on the integral rotating unit, the translation unit and the independent rotating unit after centering control is completed; inputting the stay position coordinates into a hidden Markov model for state evaluation, and obtaining a steady state probability value of the plate in the current centering control period; comparing the steady state probability value with a preset probability threshold value; When the comparison result shows that the steady state probability value is smaller than the probability threshold value, inputting the stay position coordinates and the target pose parameters into a gray prediction model for deviation trend analysis to obtain a predicted deviation value of the plate in the next centering control period; and generating a pre-compensation instruction according to the predicted deviation value, and pre-issuing the pre-compensation instruction to the servo driver so that the servo driver can perform pre-driving adjustment on the servo motors of the integral rotating unit, the translation unit and the independent rotating unit according to the pre-compensation instruction.
8. A visual-based automatic centering control system for a servo rotary table plate material is characterized by comprising the following components: the acquisition module is used for acquiring first image data of the plate in a visible light wave band and second image data of the plate in a near infrared wave band when the plate enters a preset visual field; The fusion module is used for carrying out pixel-level fusion on the first image data and the second image data to generate synthetic image data, and extracting contour point cloud information from the synthetic image data; The generating module is used for generating a digital image stream in the process of moving the plate along the conveying direction according to the definition evaluation result of the contour point cloud information; The positioning module is used for carrying out edge positioning on the digital image stream by adopting a sub-pixel edge extraction algorithm so as to obtain initial pose data of the plate, and generating target pose parameters by utilizing a camera distortion model based on the initial pose data; The comparison module is used for comparing the target pose parameter with a preset standard pose parameter based on a multi-target cooperative servo control technology to obtain a plurality of deviation values, and performing motion decoupling on the deviation values to respectively generate position compensation instructions corresponding to the integral rotation unit, the translation unit and the independent rotation unit, wherein the deviation values comprise an X-axis deviation value, a Y-axis deviation value and a rotation angle deviation value; And the control module is used for sending the position compensation instruction to a servo driver, controlling the servo motors corresponding to the integral rotating unit, the translation unit and the independent rotating unit to execute the position compensation action according to the cooperative motion path by the servo driver so as to enable the actual pose of the plate material to be consistent with the standard pose parameters and complete the centering control.
9. An electronic device, comprising: A memory for storing a computer program; a processor for implementing the steps of the vision-based servo rotary table plate automatic centering control method according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and the computer program can implement the vision-based servo rotary table plate automatic centering control method according to any one of claims 1 to 7 when executed by a processor.

Description

Automatic centering control method and system for servo rotary table plate based on vision Technical Field The application relates to the technical field of stamping automation, in particular to a vision-based automatic centering control method and system for a servo rotary table plate. Background In the field of automation of stamping production lines, the automatic centering control method of the servo rotary table plate material is a key technical link for ensuring that the plate material can be accurately placed in a die, the control precision of the method directly influences the production quality and efficiency of a subsequent stamping process, and along with continuous improvement of the requirements of automobile and household appliance manufacturing industries on the machining precision of sheet metal parts, the method has increasingly wide application prospects. The current automatic centering control method for the plate material of the servo rotary table based on vision generally adopts a single visible light camera to collect the image of the plate material, extracts the edge characteristics of the plate material through an image processing algorithm and calculates the position offset of the edge characteristics of the plate material in the transmission process, then transmits the offset to a servo control system, and drives a corresponding mechanical executing mechanism to complete the position adjustment of the plate material through a servo motor. However, in an actual production environment, when a high reflection characteristic or a low texture area exists on the surface of a plate, local overexposure or characteristic blurring easily occurs in a single visible light image, so that edge extraction accuracy is reduced, accuracy of a deviation value received by a servo control system is affected, and further the overall effect of centering adjustment is affected. Therefore, the technical problem that the accuracy of detecting the plate material level gesture needs to be further improved exists in the prior art. Disclosure of Invention The application provides a vision-based automatic centering control method and system for a servo rotary table plate, which are used for solving the problems of poor adaptability and low cooperative precision of plate centering control in the prior art. In order to solve the technical problems, in a first aspect, the application provides a vision-based automatic centering control method for a servo rotary table plate, which comprises the following steps: When a sheet material enters a preset visual field, acquiring first image data of the sheet material in a visible light wave band and second image data of the sheet material in a near infrared wave band; performing pixel-level fusion on the first image data and the second image data to generate synthetic image data, and extracting contour point cloud information from the synthetic image data; Generating a digital image stream in the process of moving the plate along the conveying direction according to the definition evaluation result of the contour point cloud information; Performing edge positioning on the digital image stream by adopting a sub-pixel edge extraction algorithm to obtain initial pose data of the plate material, and generating target pose parameters by utilizing a camera distortion model based on the initial pose data; Comparing the target pose parameter with a preset standard pose parameter based on a multi-target cooperative servo control technology to obtain a plurality of deviation values, and performing motion decoupling on the deviation values to respectively generate position compensation instructions corresponding to an integral rotation unit, a translation unit and an independent rotation unit, wherein the deviation values comprise an X-axis deviation value, a Y-axis deviation value and a rotation angle deviation value; And the position compensation instruction is issued to a servo driver, and the servo driver controls the servo motors corresponding to the integral rotating unit, the translation unit and the independent rotating unit to execute position compensation according to the cooperative motion path, so that the actual pose of the plate material is consistent with the standard pose parameter, and centering control is completed. In a second aspect, the application provides a vision-based automatic centering control system for a servo rotary table plate, comprising: the acquisition module is used for acquiring first image data of the plate in a visible light wave band and second image data of the plate in a near infrared wave band when the plate enters a preset visual field; The fusion module is used for carrying out pixel-level fusion on the first image data and the second image data to generate synthetic image data, and extracting contour point cloud information from the synthetic image data; The generating module is used for generating a digital image stream in the process of m