A grabbing deviation rectification method and system based on visual detection technology

Abstract

This invention relates to the field of image processing technology, and more particularly to a grasping and correction method and system based on visual detection technology. The method includes: acquiring images of all workpieces to be grasped after unloading; for any given image, determining the center coordinates and deflection angle relative to the horizontal direction of the corresponding workpiece, and then storing them in a correction list; the center coordinates and deflection angle of each workpiece to be grasped constitute a data entry in the correction list; retrieving a data entry from the correction list and sending it to a grasping robot to grasp the workpiece corresponding to that data entry; in response to the completion of grasping, deleting that data entry from the correction list and retrieving the next data entry to send to the grasping robot, until the correction list is empty. The method of this invention can successfully grasp workpieces.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a grasping and correction method and system based on visual detection technology. Background Technology

[0002] When processing sheet metal, a single sheet is typically cut into multiple workpieces that meet dimensional requirements. During the cutting process, a cutting tool creates cutting paths on the sheet, dividing the originally whole sheet into several independent workpieces. After cutting, to enable subsequent processing, transfer, or sorting operations, these workpieces need to be moved from the cutting location to a designated unloading position. This is usually done by using a pusher to move all workpieces as a whole. However, because the workpieces are independent, the friction and thrust transmission effects experienced by each workpiece differ during the pusher's movement, causing varying degrees of positional shift in each workpiece.

[0003] When a robot is grasping a workpiece, if it still controls the suction cup to move to the corresponding position to perform the grasping action according to the preset positioning coordinates (such as the center coordinates of each workpiece before unloading obtained by vision or mechanical positioning), the grasping will be off-center or miss, resulting in the inability to successfully grasp the workpiece.

[0004] Therefore, how to avoid missing or misgrabbing when there is positional deviation is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] To address the technical problem of misalignment or missed grasping when the workpiece is misaligned, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a grasping correction method based on visual inspection technology, comprising: acquiring images of all workpieces to be grasped after unloading; for any image, determining the center coordinates of the workpiece to be grasped corresponding to the image and the deflection angle relative to the horizontal direction, and then storing them in a correction list; the center coordinates and deflection angle of each workpiece to be grasped constitute a data entry in the correction list; retrieving a data entry from the correction list and sending it to a grasping robot so that the grasping robot grasps the workpiece to be grasped corresponding to the data entry; in response to the completion of grasping, deleting the data entry from the correction list and retrieving the next data entry and sending it to the grasping robot, until the correction list is empty.

[0007] Further, determining the center coordinates of the workpiece to be grasped corresponding to the image includes: obtaining the theoretical center coordinates of the workpiece to be grasped; recognizing the image to obtain the outline of the corresponding workpiece to be grasped; calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped; and determining the center coordinates based on the theoretical center coordinates and the actual center coordinates; wherein the center coordinates are:

[0008]

[0009] In the formula, (X C ,Y C (X) represents the center coordinates of the Cth workpiece to be grasped, K is the uniform scaling factor, and θ is the deflection angle. Offsetc ,Y Offsetc Let t be the actual center coordinates of the Cth workpiece to be grasped. x The first offset represents the horizontal translation amount; t y The second offset represents the amount of translation in the vertical direction.

[0010] Furthermore, the outline is a polygon; calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped includes: calculating the centroid of the outline of the corresponding workpiece to be grasped, and using the centroid as the actual center coordinates.

[0011] Furthermore, the expression for calculating the deflection angle is:

[0012]

[0013] In the formula, θ is the deflection angle, and t x For the first offset, t y The second offset is obtained by subtracting the abscissa of the actual center coordinate from the abscissa of the theoretical center coordinate; the second offset is obtained by subtracting the ordinate of the actual center coordinate from the ordinate of the theoretical center coordinate.

[0014] Furthermore, it also includes: determining whether the identified contour matches a certain contour in the workpiece to be grasped; if not, determining the similarity between the identified contour and the contours in all the workpieces to be grasped; determining whether the maximum similarity among all similarities is greater than a preset similarity threshold; if so, taking the contour corresponding to the maximum similarity as the identified contour.

[0015] Furthermore, before grasping, the process includes: calculating the distance between the actual center coordinates and the theoretical center coordinates, determining whether the distance is greater than a preset distance threshold, and if so, controlling the grasping robot to grasp based on the deflection angle and the center coordinates; otherwise, grasping at the originally set position.

[0016] Furthermore, before calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped, the grasping robot grasps the workpiece using a suction cup, and before calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped, the robot further includes: calculating the weight of the workpiece to be grasped based on the area and weight per unit area of ​​the workpiece to be grasped, and determining whether the weight is greater than the maximum load capacity of the current suction cup. If so, an alarm is issued.

[0017] Furthermore, it also includes: acquiring images via two vision cameras on the suction cup.

[0018] Furthermore, the suction cup is a rectangular suction cup.

[0019] In a second aspect, the present invention provides a grasping and correction system based on visual inspection technology, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the grasping and correction method based on visual inspection technology described in the first aspect is implemented.

[0020] The beneficial effects of this invention are as follows: by controlling the gripping robot to grip the workpiece according to the center coordinates and deflection angle of the workpiece after unloading, it is possible to avoid gripping off-center or missing the workpiece due to gripping according to the position and angle before correction, thereby improving gripping efficiency; by retrieving data from the correction list for gripping while calculating the center coordinates and deflection angle of the workpiece, gripping efficiency can be improved; furthermore, by performing correction only when the distance between the actual center coordinates and the theoretical center coordinates is greater than a preset distance threshold, invalid correction operations can be reduced, thereby improving gripping efficiency. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a grasping and correction method based on visual detection technology according to an embodiment of the present invention;

[0022] Figure 2 This is a flowchart illustrating the determination of the center coordinates and deflection angle of the workpiece to be grasped;

[0023] Figure 3 This is a schematic diagram illustrating the structure of a grasping and correction system based on visual detection technology according to an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] Figure 1 This is a flowchart illustrating a grasping and correction method based on visual detection technology according to an embodiment of the present invention.

[0027] To address the technical problem of workpieces shifting when pushed to the unloading position, preventing successful grasping by the gripping robot, this invention, in a first aspect, provides a gripping correction method based on visual detection technology. Specifically, as... Figure 1 As shown, the method of the present invention includes the following steps.

[0028] S1. Collect images of all workpieces to be grabbed after unloading.

[0029] Specifically, the gripping robot is controlled to reach a designated location (multiple preset photo points) to capture images of all workpieces to be gripped (i.e., workpieces obtained after cutting a whole piece of sheet metal). In this embodiment, images are captured by two vision cameras mounted on the gripping robot's suction cup, one of which is used to capture the depth of the workpiece to identify all workpieces, such as partially covered workpieces. Furthermore, the suction cup referred to is a rectangular suction cup; in other optional embodiments, it can also be a circular suction cup.

[0030] S2. For any image, determine the center coordinates of the workpiece to be grasped corresponding to the image, as well as the deflection angle relative to the horizontal direction, and then store it in the correction list.

[0031] Specifically, such as Figure 2 As shown, determining the center coordinates and deflection angle of the workpiece to be grasped includes the following steps.

[0032] S201. Obtain the theoretical center coordinates of the workpiece to be grabbed.

[0033] In this embodiment, the theoretical center coordinates are the coordinates of the center points of each workpiece on the board that has been laid out before cutting (pre-stored in the system). In other optional embodiments, the theoretical center coordinates can also be the coordinates of the center point of the workpiece to be picked up at the expected destination position.

[0034] S202. Recognize the image to obtain the outline of the corresponding workpiece to be grasped.

[0035] Because the images captured by the vision camera are too large and contain multiple workpieces (polygonal plates, such as rectangular plates), it is necessary to first extract the contours of each workpiece in the image. Specifically, the captured image is preprocessed, such as by converting it to grayscale and enhancing its contrast. Then, edge detection is performed on the preprocessed image to obtain the edge detection results (binarized edge map), and a contour detection algorithm is used to extract the contours from the edge detection results (binarized edge map). In this embodiment, the Canny algorithm is used for edge detection. Since the specific process is existing technology, it will not be described in detail here.

[0036] In an optional embodiment, the method of the present invention may further include: determining whether the extracted contour matches a certain contour in the workpiece to be grasped, that is, comparing the extracted contour with the length and width of the contours of all the workpieces to be grasped (the contours, lengths, and widths of each workpiece to be grasped are stored in the system). If they match, it is a match; if they do not match, it is a mismatch. If they do not match, it may be because the two workpieces are partially stacked, or the complete workpiece was not captured, resulting in an incomplete or inaccurate extracted contour. Therefore, the similarity between the extracted contour and the contours of all the workpieces to be grasped can be calculated. If the maximum similarity among all the obtained similarities is greater than a preset similarity threshold (set according to actual needs), then the contour corresponding to the maximum similarity is taken as the extracted contour to obtain a complete and accurate contour; if the similarity is less than or equal to the preset similarity threshold, it indicates that the accuracy of the identified contour is low, and the identified contour is discarded to avoid invalid grasping operations based on the contour.

[0037] In this embodiment, the similarity between two contours can be calculated using Hausdorff distance or Euclidean distance.

[0038] For incomplete or inaccurate contours, a highly similar contour is searched in a pre-set database and replaced with the incomplete or inaccurate contour. This improves the reliability of obtaining the workpiece contour, thereby increasing the success rate of subsequent grasping. Calculating the center coordinates and deflection angle by referring to the image before cutting (i.e., the image acquired when obtaining the theoretical center coordinates of the workpiece) reduces processing difficulty compared to directly calculating from the image after cutting. This means that the workpiece can be searched purposefully (within possible angles) based on the image before cutting, avoiding the time wasted by blindly searching for the workpiece.

[0039] This yields the contours of each workpiece.

[0040] S203. Calculate the actual center coordinates of the contour of the corresponding workpiece to be grasped.

[0041] For any workpiece to be gripped (denoted as the target workpiece), calculate the centroid of the target workpiece's contour, and then use this centroid as the actual center coordinates of the target workpiece (the coordinates of the center point of the workpiece after unloading). Specifically, obtain the coordinates of each vertex of the target workpiece's contour; calculate the area of ​​the target workpiece's contour based on the coordinates of each vertex. In this embodiment, the expression for calculating the area of ​​the target workpiece's contour is:

[0042]

[0043] In the formula, S is the area of ​​the contour of the target workpiece, and x i Let y be the x-coordinate of the i-th vertex of the target workpiece. i Let x be the ordinate of the i-th vertex of the target workpiece. i+1 Let y be the x-coordinate of the (i+1)th vertex of the target workpiece. i+1 Let be the ordinate of the (i+1)th vertex of the target workpiece.

[0044] Furthermore, the centroid of the target workpiece's contour is calculated based on the area of ​​the contour and the coordinates of each vertex. Specifically, the expression for calculating the centroid (X_c, Y_c) of the target workpiece's contour is:

[0045]

[0046] In the formula, S is the area of ​​the contour of the target workpiece, and x i Let y be the x-coordinate of the i-th vertex of the target workpiece. i Let x be the ordinate of the i-th vertex of the target workpiece. i+1 Let y be the x-coordinate of the (i+1)th vertex of the target workpiece. i+1 Let be the ordinate of the (i+1)th vertex of the target workpiece.

[0047] This yields the actual center coordinates of the target workpiece. It should be noted that these actual center coordinates and the theoretical center coordinates belong to the same coordinate system.

[0048] S204. Determine the center coordinates and deflection angle of the target workpiece based on the theoretical center coordinates and the actual center coordinates.

[0049] Specifically, the first offset is obtained by subtracting the abscissa of the actual center coordinate of the target workpiece from the abscissa of the theoretical center coordinate. This first offset represents the difference between the actual center coordinate and the theoretical center coordinate in the horizontal direction (i.e., the X direction). The second offset is obtained by subtracting the ordinate of the actual center coordinate of the target workpiece from the ordinate of the theoretical center coordinate. This second offset represents the difference between the actual center coordinate and the theoretical center coordinate in the vertical direction (i.e., the Y direction).

[0050] Further, the deflection angle is calculated based on the first offset and the second offset. This deflection angle represents the angular deviation in the horizontal direction between the actual center coordinates and the theoretical center coordinates. In one embodiment, the expression for calculating the deflection angle is:

[0051]

[0052] In the formula, θ is the deflection angle, and t x For the first offset, t y This is the second offset.

[0053] Furthermore, by performing a graphic similarity transformation on the actual center coordinates and adding a uniform scaling factor K, the center coordinates of the workpiece's coordinate system can be obtained. In one embodiment, the expression for calculating the center coordinates of the target workpiece is:

[0054]

[0055] That is:

[0056]

[0057] In the formula, (X C ,Y C (X) represents the center coordinates of the Cth workpiece to be grasped, K is the uniform scaling factor, and θ is the deflection angle. Offsetc ,Y Offsetc Let t be the actual center coordinates of the Cth workpiece to be grasped. x t is the first offset; y This is the second offset.

[0058] Furthermore, the deflection angle and center coordinates of the target workpiece are stored in the correction list, and the center coordinates and deflection angle of each workpiece to be grasped constitute a data entry in the correction list.

[0059] S3. Take a data point from the correction list and send it to the gripping robot so that the gripping robot can grip the workpiece corresponding to that data point.

[0060] Specifically, a data point is retrieved from the correction list and sent to the gripping robot. Upon receiving the data, the gripping robot moves to align the center coordinates of the suction cup with the center coordinates of the workpiece to be gripped, and controls the suction cup to rotate by the corresponding angle based on the deflection angle, so that the suction cup can successfully grip the target workpiece. Further, in response to the completion of gripping, the data point is deleted from the correction list, and the next data point is retrieved and sent to the gripping robot. The above process is repeated until the correction list is empty.

[0061] It should be noted that calculating the center coordinates and offset angle of the workpiece and retrieving data from the correction list and sending it to the gripping robot are performed in parallel; that is, S2 and S3 are performed simultaneously. This simultaneous calculation and retrieval improves gripping efficiency. Furthermore, compared to first acquiring an image of a workpiece at a pre-set position, then calculating its center coordinates and offset angle, gripping that workpiece based on these coordinates, and then acquiring the next workpiece image after gripping, this invention achieves continuous gripping without interruption by first capturing images of all workpieces and then performing calculations and gripping simultaneously, further improving gripping efficiency.

[0062] By controlling the gripping robot to grasp the workpiece according to the center coordinates and deflection angle of the workpiece after unloading, it is possible to avoid grasping off-center or missing the workpiece due to grasping according to the position and angle before correction, thereby improving the gripping efficiency.

[0063] In one embodiment, before grasping, the method of the present invention further includes: calculating the distance (Euclidean distance) between the actual center coordinates and the theoretical center coordinates of the target workpiece, and determining whether the distance is greater than a preset distance threshold (set to 3cm in this embodiment). If so, it indicates that the difference between the actual center coordinates and the theoretical center coordinates is large and correction is required; otherwise, it is easy to miss or deviate from the target, resulting in failure to grasp. If not, it indicates that the difference between the actual center coordinates and the theoretical center coordinates is small, and the workpiece can be grasped successfully without position and angle correction. In this case, the grasp can be performed directly according to the originally set position and rotation angle.

[0064] Since the distance between the actual center coordinates and the theoretical center coordinates is small, there is generally no situation where the capture is off-center or misses the target. Therefore, there is no need to perform correction operations, thereby reducing invalid correction operations and improving the capture efficiency.

[0065] In an optional embodiment, the method of the present invention further includes: when the distance between the actual center coordinates and the theoretical center coordinates of the target workpiece is less than a distance threshold, if direct grasping fails, then the deflection angle and center coordinates of the workpiece are calculated, and then grasping is performed based on the deflection angle and center coordinates (i.e., correction is performed).

[0066] By performing corrective gripping after a direct gripping failure (without correction), the success rate of gripping can be further improved; compared to manually moving the workpiece for correction, the gripping efficiency is improved.

[0067] In one embodiment, the method of the present invention further includes: calculating the weight of the target workpiece based on the area and weight per unit area of ​​the target workpiece, determining whether the weight exceeds the maximum load capacity of the current suction cup, and issuing an alarm if so. The weight per unit area needs to be measured in advance, and the maximum load capacity of the suction cup is pre-stored in the system.

[0068] By issuing an alarm when the weight of the workpiece exceeds the maximum load capacity of the suction cup, the system can prompt relevant personnel to replace the suction cup or increase the number of suction cups to avoid situations where the current suction cup cannot successfully grip the workpiece.

[0069] In an optional embodiment, the method of the present invention may further include: obtaining the working area of ​​the suction cup corresponding to the actual center point coordinates (the working area of ​​the suction cup on the workpiece, i.e., the adsorption area, can be obtained based on the coincidence of the center of the suction cup and the center of the workpiece); dividing the working area into multiple detection areas (based on the area that can affect the successful adsorption as a region, for example, if an impurity or depression of area A will affect the gripping, the number of regions can be A divided by the area of ​​the working area of ​​the suction cup); for any detection area, calculating the mean gray value of the pixels in the detection area, and determining whether the mean value is greater than a preset gray value threshold (i.e., the surface is free of impurities or defects, or even if the surface has impurities or defects). If the average grayscale value of the pixels in the image corresponding to the workpiece being gripped by the suction cup is less than or equal to the average grayscale value of the workpiece (the workpiece and the target workpiece come from the same sheet material), then the area to be detected is considered an abnormal area. Simultaneously, the proportion of abnormal pixels in this abnormal area to the total number of pixels in the abnormal area is calculated. Abnormal pixels are defined as pixels with grayscale values ​​greater than a grayscale value threshold. If this proportion is greater than a set threshold (e.g., 70%), it indicates that the abnormal area may affect gripping. The working area of ​​the suction cup is then adjusted so that the workpiece area of ​​the suction cup does not include the abnormal area. In other words, the working area of ​​the suction cup will only be adjusted when the area is an abnormal area and the proportion corresponding to the abnormal area is greater than the set threshold.

[0070] When surface dents and impurities on the workpiece affect gripping, adjusting the working area of ​​the suction cup to eliminate abnormal areas (containing impurities or dents) further improves the success rate of gripping. It should be noted that if the adjusted working area cannot grip the workpiece, then no adjustment is made. For example, if the workpiece size matches the suction cup size, even with surface dents or impurities, no adjustment is necessary because adjustments might result in some areas of the suction cup not adhering to the workpiece (gap areas / gap areas, especially large ones, preventing successful gripping).

[0071] In this embodiment, the specific process of using a suction cup to grasp a workpiece is as follows: The device is turned on, the coordinates of the suction cup's movement are sent to the grasping robot, and the suction cup's status is read to determine if it has moved into position. If so, the camera deployed on the suction cup takes a picture. Further, it is determined whether the entire sheet material has been photographed (i.e., whether images of all workpieces to be grasped have been captured). If so, all captured images are fed back to the device's processing module to calculate the center coordinates and deflection angle of the workpiece, and the calculated center coordinates and deflection angle are stored in a correction list. Simultaneously, it is determined whether there is a data column in the correction list. If so, a data entry is taken from the correction list and sent to the grasping robot so that the robot can grasp the workpiece based on that data. After the workpiece grasping is completed, it is determined whether the process is complete. If complete, the data entry is deleted from the correction list, and the next data entry is retrieved from the correction list, and so on, until the correction list is empty.

[0072] Figure 3 This is a schematic diagram illustrating the structural block diagram of the grasping and correction system based on visual detection technology according to this embodiment.

[0073] In a second aspect, the present invention also provides a grasping and correction system based on visual detection technology. For example... Figure 3 As shown, the grasping and correction system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a grasping and correction method based on visual detection technology according to the first aspect of the present invention.

[0074] The grasping and correction system also includes other components well known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0075] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.

[0076] In the description of this specification, "multiple" means at least two, such as two, three or more, unless otherwise explicitly specified. Furthermore, the steps described above are for clarity only; in implementation, they can be combined into one step or some steps can be broken down into multiple steps, as long as they include the same logical relationships.

[0077] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A grasping and correction method based on visual detection technology, characterized in that, include: Collect images of all workpieces to be grabbed after unloading; For any image, determine the center coordinates of the workpiece to be grasped corresponding to the image, as well as the deflection angle relative to the horizontal direction, and then store them in the correction list; the center coordinates and deflection angle of each workpiece to be grasped constitute a data entry in the correction list; at the same time, take a data entry from the correction list and send it to the grasping robot so that the grasping robot can grasp the workpiece to be grasped corresponding to that data entry; the grasping robot grasps the workpiece using a suction cup, which is rectangular; During the gripping process, the working area of ​​the suction cup corresponding to the actual center coordinates is obtained, and this working area is divided into multiple detection areas. For any detection area, the average gray value of the pixels in the detection area is calculated, and it is determined whether the average value is greater than a preset gray value threshold. If so, the detection area is regarded as an abnormal area, and the proportion of the number of abnormal pixels in the abnormal area to the total number of pixels in the abnormal area is calculated. The abnormal pixels are those with gray values ​​greater than the gray value threshold. If the proportion is greater than the set threshold, the working area of ​​the suction cup is adjusted so that the workpiece area of ​​the suction cup does not include the abnormal area. In response to the completion of the capture, the data is removed from the correction list and the next data is sent to the capture robot until the correction list is empty; Determining the center coordinates of the workpiece to be grasped corresponding to the image includes: obtaining the theoretical center coordinates of the workpiece to be grasped; recognizing the image to obtain the outline of the corresponding workpiece to be grasped; calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped; and determining the center coordinates based on the theoretical center coordinates and the actual center coordinates; wherein the center coordinates are: ； In the formula, ( ) is the first The center coordinates of the workpiece to be grabbed The scaling factor is uniform. For the deflection angle, ( ) is the first The actual center coordinates of the workpiece to be grabbed. The first offset represents the amount of translation in the horizontal direction; The second offset represents the amount of translation in the vertical direction; The outline is a polygon; calculating the actual center coordinates of the outline of the corresponding workpiece to be grasped includes: calculating the centroid of the outline of the corresponding workpiece to be grasped, and using the centroid as the actual center coordinates.

2. The grasping and correction method based on visual detection technology according to claim 1, characterized in that, The expression for calculating the deflection angle is: ； In the formula, For the deflection angle, This is the first offset. The second offset is obtained by subtracting the abscissa of the actual center coordinate from the abscissa of the theoretical center coordinate; the second offset is obtained by subtracting the ordinate of the actual center coordinate from the ordinate of the theoretical center coordinate.

3. The grasping and correction method based on visual detection technology according to claim 1, characterized in that, Also includes: Determine whether the identified contour matches a certain contour in the workpiece to be grasped. If not, determine the similarity between the identified contour and the contours in all the workpieces to be grasped. Determine whether the maximum similarity among all similarities is greater than a preset similarity threshold. If so, take the contour corresponding to the maximum similarity as the identified contour.

4. The grasping and correction method based on visual detection technology according to claim 1, characterized in that, Before grasping, the process also includes: calculating the distance between the actual center coordinates and the theoretical center coordinates, determining whether the distance is greater than a preset distance threshold, and if so, controlling the grasping robot to grasp based on the deflection angle and the center coordinates; otherwise, grasping at the originally set position.

5. The grasping and correction method based on visual detection technology according to claim 1, characterized in that, Before calculating the actual center coordinates of the corresponding workpiece outline, the process also includes: calculating the weight of the workpiece based on its area and weight per unit area, determining whether the weight is greater than the maximum load capacity of the current suction cup, and issuing an alarm if so.

6. The grasping and correction method based on visual detection technology according to claim 1, characterized in that, Also includes: Images are captured using two vision cameras mounted on a suction cup.

7. A grasping and correction system based on visual detection technology, characterized in that, It includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the grasping and correction method based on visual detection technology according to any one of claims 1-6 is implemented.

Images