A machine vision-based suction cup deviation correction method and system

Abstract

The present application relates to the technical field of image processing, in particular to a kind of based on machine vision's suction cup deviation rectification method and system.The method includes: according to the preset coordinate of stored target workpiece, move suction cup to preset position;Image of the target workpiece is collected, the center coordinates of the outline of the target workpiece are identified, and the center coordinates of the outline are converted target workpiece real center coordinates;The offset distance of preset coordinate and real center coordinates of target workpiece is obtained, if offset distance is less than threshold value, then directly carry out workpiece's grabbing;If offset distance is greater than or equal to threshold value, then calculate deviation rectification index, if deviation rectification index is greater than set threshold value, then the preset position of suction cup is rectified;Otherwise, directly carry out workpiece's grabbing;Deviation rectification index represents the situation that suction cup deviates from.It is that the scheme of the present application can more efficiently, reliably carry out deviation rectification when suction cup grabs workpiece.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a suction cup correction method and system based on machine vision. Background Technology

[0002] In the furniture manufacturing industry, the use of suction cups for workpiece gripping is a key link in the automated production process. That is, suction cup gripping can be combined with automated equipment (such as robotic arms and conveyor belts) to realize the automatic transfer of workpieces between processes such as cutting, edge banding, drilling, and assembly without human intervention. For example, after cutting, the cut workpiece needs to be pushed to the unloading position, and the board is transferred from the cutting machine to the edge banding machine by the suction cup.

[0003] The suction cup gripping relies on preset positioning coordinates (such as the center coordinates of each workpiece before unloading, obtained through vision or mechanical positioning) and the flatness of the workpiece surface.

[0004] For example, after the material is cut, there will be tool marks between the workpieces. When the machine pushes the workpieces as a whole, if the pushing force is uneven, there is friction between the workpieces, or the material (such as wood) itself has slight deformation, it is easy to cause the workpieces to shift in position (such as left-right misalignment, forward-backward overtravel) or tilt in posture (such as tilting, rotation) after being unloaded. That is, when the workpieces are unloaded to the unloading position, each workpiece will have different degrees of deviation and displacement. If the actual position of the workpiece deviates from the preset deviation from the suction cup's tolerance range, it will directly cause the suction cup to "grab empty" or "grab off-center," or only partially contact the workpiece, failing to form effective adsorption.

[0005] For example, if there are pits on the workpiece during the cutting process (which are "hard defects" on the workpiece surface), it will directly disrupt the tight fit between the suction cup and the workpiece surface, causing the vacuum seal to fail or the adsorption force to decrease during the adsorption process, resulting in unstable adsorption.

[0006] Therefore, when a suction cup grips a workpiece, it is particularly important to perform positioning and correction to enable more efficient and safer automatic transfer of the workpiece. Summary of the Invention

[0007] The purpose of this invention is to propose a suction cup correction method and system based on machine vision, in order to solve the problem in the prior art that the suction cup cannot efficiently and safely perform workpiece positioning and correction, resulting in low automatic transfer efficiency; to this end, this invention provides solutions in the following two aspects.

[0008] In a first aspect, the present invention provides a suction cup correction method based on machine vision, comprising:

[0009] Based on the preset coordinates of the target workpiece stored in the database, move the suction cup to the preset position;

[0010] The image of the target workpiece is acquired, the center coordinates of the outline of the target workpiece are identified, and the center coordinates of the outline are converted into the true center coordinates of the target workpiece.

[0011] The offset distance between the preset coordinates of the target workpiece and the actual center coordinates is obtained. If the offset distance is less than a threshold, the workpiece is directly grasped. If the offset distance is greater than or equal to the threshold, a correction index is calculated. If the correction index is greater than the set threshold, the preset position of the suction cup is corrected. Otherwise, the workpiece is directly grasped. The correction index represents the deviation of the suction cup.

[0012] The above solution estimates the relative position of the target workpiece and the suction cup by judging the offset distance between the preset coordinates and the center coordinates. For those with little change in relative position, the suction cup can be used directly to grasp them. For those with large change in relative position, it is necessary to use the correction index to further determine whether the suction cup has moved. This allows for more efficient and reliable correction when the suction cup grasps the workpiece.

[0013] Optionally, the suction cup is circular or rectangular in shape. When the suction cup is rectangular, the correction index... for:

[0014] ;

[0015] in, Let c be the degree of overlap of the c-th workpiece. Let c be the mass percentage of the c-th workpiece. It is an exponential function with the natural number e as the base. , Let be the length and width of the c-th workpiece, respectively. , These are the length and width of the suction cup, respectively; the degree of overlap is the ratio of the area of ​​the suction cup to the area of ​​the workpiece region in the projected area of ​​the c-th workpiece; and the mass ratio is the ratio of the weight of the workpiece to the maximum weight that the suction cup can bear.

[0016] The aforementioned correction index determines whether the suction cup needs correction in two ways. First, when the size of the workpiece is greater than or equal to the size of the suction cup, the degree of overlap and the mass ratio are considered to accurately assess whether the suction cup needs correction. Second, when the size of the workpiece does not meet the requirement of being greater than or equal to the size of the suction cup, it is directly determined that correction is required.

[0017] Optionally, converting the center coordinates of the contour to the true center coordinates of the target workpiece involves converting pixel coordinates to the world coordinate system.

[0018] Optionally, correcting the preset position of the suction cup includes:

[0019] Calculate the offset angle between the preset coordinates and the actual center coordinates;

[0020] The suction cup is rotated according to the offset angle.

[0021] The above method improves the positioning efficiency of the suction cup by simply adjusting its angle.

[0022] Optionally, the true center coordinates are:

[0023] ;

[0024] in,( , () represents the center coordinates of the profile of the c-th workpiece. , Let be the true center coordinates of the c-th workpiece after blanking, and K be the uniform scaling factor. Let be the relative rotation angle between the line connecting the center coordinates of the workpiece's contour and the theoretical center coordinates and the x-direction. , These are the x-direction translation vectors and y-direction translation vectors of the center coordinates of the contour before and after loading, relative to the theoretical center coordinates. , At least one of them is not 0; the theoretical center coordinates are the center coordinates of the workpiece image on the wooden board that has been laid out before cutting, and the wooden board is located on the xOy plane.

[0025] Optionally, the relative rotation angle for:

[0026] ;

[0027] in, , These are the x-direction translation vector and y-direction translation vector, respectively, between the center coordinates of the workpiece's contour in the image before and after loading and the theoretical center coordinates. It is an inverse cosine function.

[0028] Optionally, before gripping the workpiece, the following steps are also included:

[0029] like Then, the region image corresponding to the suction cup is acquired, and the region image is processed into grayscale to obtain a grayscale image; , Let be the length and width of the c-th workpiece, respectively. , These are the length and width of the suction cup, respectively.

[0030] Obtain the difference image between the grayscale image and the standard image, perform thresholding on the difference image, and extract abnormal regions;

[0031] If the area of ​​the abnormal region exceeds the set value, the position of the suction cup will be readjusted.

[0032] The above solution mitigates the impact by readjusting the suction cup gripping position (avoiding defective areas, such as pits or areas with rough textures); otherwise, it may easily cause production failures.

[0033] Alternatively, the image is acquired by a camera deployed on a suction cup.

[0034] Optionally, it may also include a step of performing image enhancement processing on the image.

[0035] In the second aspect, a machine vision-based suction cup correction system includes:

[0036] processor;

[0037] The memory stores computer instructions for a machine vision-based suction cup alignment method, which, when executed by the processor, causes the system to perform the aforementioned machine vision-based suction cup alignment method.

[0038] The beneficial effects of this invention are as follows:

[0039] The present invention first performs a preliminary analysis by comparing the distance between the preset coordinates of the target workpiece and the actual center coordinates of the target workpiece after unloading. This analysis determines whether the target workpiece needs to be corrected. It can determine if the target workpiece has shifted. If the shift is small, no correction by the suction cup is needed. If the shift is large, further correction is required, combining the degree of overlap between the suction cup area and the target workpiece area, as well as the weight of the workpiece, to determine whether correction is necessary. Through this method, the present invention achieves greater efficiency and accuracy in workpiece transfer. Attached Figure Description

[0040] Figure 1 The flowchart illustrating the steps of a suction cup correction method based on machine vision in this embodiment is shown in the schematic diagram.

[0041] Figure 2 The diagram illustrates the structure of a machine vision-based suction cup correction system in this embodiment. Detailed Implementation

[0042] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0043] This invention targets workpieces of different sizes after the sheet metal has been cut. After the sheet metal is cut, there will be tool paths between the workpieces. When the machine pushes the workpieces out as a whole, the workpieces will have different degrees of deviation and displacement. However, the robot still grabs the workpieces according to the coordinates set before the cutting. At this time, the grab will be off-center.

[0044] To address the aforementioned problems, this invention proposes a suction cup correction method based on machine vision.

[0045] Specifically, such as Figure 1 As shown, a suction cup correction method based on machine vision in this embodiment includes the following steps:

[0046] Step S1: Move the suction cup to the preset position according to the preset coordinates of the stored target workpiece.

[0047] Specifically, the preset coordinates of the target workpiece are the center coordinates (X, Y, Z) of each workpiece on the pre-laid wooden board before cutting. 0c ,Y 0c ,0), where c represents the serial number of the target workpiece.

[0048] It should be noted that there are multiple workpieces of different sizes on the pre-laid wooden boards before cutting, and each workpiece corresponds to a preset coordinate.

[0049] Since the suction cup is used for workpiece transfer, theoretically, the suction cup needs to be moved directly above the target workpiece first. Therefore, after obtaining the preset coordinates of the target workpiece, the robot can convert them into the preset position of the suction cup and then determine the position based on the coordinates of the suction cup's preset position, i.e., (X...). 0c ,Y 0c Z 0c ), control the suction cup to move to the preset position.

[0050] The coordinate transformation mentioned above can be a transformation between the image coordinate system and the camera coordinate system. Since the specific transformation process is existing technology, it will not be described in detail here.

[0051] Step S2: Acquire an image of the target workpiece, identify the center coordinates of the outline of the target workpiece, and convert the center coordinates of the outline into the true center coordinates.

[0052] In this embodiment, after the robot controls the suction cup to move to a preset position, images are acquired through a vision acquisition camera on the suction cup.

[0053] Since the images captured by the vision acquisition camera are too large and may contain multiple workpieces, it is necessary to first extract the outline of the target workpiece in the image, and then obtain the center coordinates of the outline of the target workpiece.

[0054] In one embodiment, the process of extracting the contour of the target workpiece is as follows:

[0055] Edge detection is performed on the image to obtain the edge detection result (binarized edge map), and a contour detection algorithm is used to extract the contour from the edge detection result (binarized edge map);

[0056] By comparing the aspect ratio of each workpiece in the image with the actual aspect ratio of the target workpiece, if they are consistent, the corresponding contour is taken as the contour of the target workpiece.

[0057] The edge detection described above uses the Canny algorithm. Since the specific process is existing technology, it will not be elaborated upon here.

[0058] The above method can accurately identify the target workpiece and obtain the center coordinates of the contour of the target workpiece that may be offset.

[0059] In one embodiment, since the center coordinates of the contour are the coordinates of the target workpiece in the pixel coordinate system, while the true center coordinates of the workpiece are in the world coordinate system, it is necessary to transform the center coordinates in the pixel coordinate system to the world coordinate system in order to obtain the true center coordinates of the workpiece.

[0060] Since the conversion from pixel coordinates to world coordinates is based on existing technology, it will not be elaborated here.

[0061] In another embodiment, the process of obtaining the true center coordinates of the workpiece after blanking is as follows:

[0062] First, obtain the x-direction translation vector and y-direction translation vector of the center coordinates of the contour before and after loading, and the theoretical center coordinates.

[0063] Specifically, the theoretical center coordinates are the center coordinates of the workpiece image on the pre-laid wooden board before cutting, which correspond to the preset coordinates, and the wooden board is located on the xOy plane.

[0064] It should be noted that when , When both are 0, it means there is no translational deviation, i.e., the workpiece does not shift, and no angle correction is needed; that is, θ can be considered 0. When , If at least one value is not 0, it means that the workpiece has shifted, and the true center coordinates should be obtained at this time.

[0065] Secondly, when the workpiece shifts, the center coordinates of the contour are transformed into the true center coordinates through image similarity transformation.

[0066] Specifically, the transformation formula for the true center coordinates is as follows:

[0067] ;

[0068] get:

[0069] ;

[0070] .

[0071] in,( , () represents the center coordinates of the profile of the c-th workpiece. , Let be the true center coordinates of the c-th workpiece after blanking, and K be the uniform scaling factor. Let be the relative rotation angle between the line connecting the center coordinates of the workpiece's contour and the theoretical center coordinates and the x-direction. , These are the x-direction translation vectors and y-direction translation vectors of the center coordinates of the contour before and after loading, and the theoretical center coordinates, respectively.

[0072] Among them, relative rotation angle for:

[0073] ;

[0074] in, , These are the x-direction translation vector and y-direction translation vector, respectively, between the center coordinates of the workpiece's contour in the image before and after loading and the theoretical center coordinates. It is an inverse cosine function.

[0075] During the coordinate transformation described above, the pixel coordinates of the known workpiece are used to obtain the true center coordinates of the coordinate system containing the actual workpiece by adding a uniform scaling factor K through graphic similarity transformation. , ).

[0076] It should be noted that when capturing images of the workpiece, the camera and the plane containing the workpiece are parallel. This means that the pixel coordinates and world coordinates in the image captured by the camera are both on the xOy plane. In this case, theoretically, the offset and deflection between the theoretical center coordinates of the workpiece and the center coordinates of the contour in the pixel coordinate system can also be reflected in the coordinates of the real workpiece. Therefore, the above transformation formula takes into account the parallel nature of the plane containing the camera and the workpiece, and directly translates, rotates, and scales the center coordinates of the contour of the target workpiece to obtain the real center coordinates of the workpiece. Compared with converting pixel coordinates to world coordinates, this method is simpler.

[0077] Step S3: Obtain the preset coordinates of the target workpiece and the offset distance of the actual center coordinates. If the offset distance is less than the threshold, the workpiece is directly grasped. If the offset distance is greater than or equal to the threshold, the correction index is calculated. If the correction index is greater than the set threshold, the preset position of the suction cup is corrected. Otherwise, the workpiece is directly grasped.

[0078] The aforementioned offset distance represents the deviation of the target workpiece before and after unloading, i.e., the deviation resulting from the change in the center coordinates of the workpiece before and after unloading.

[0079] The offset of the target workpiece is obtained by calculating the Euclidean distance between the preset coordinates and the actual center coordinates. This Euclidean distance calculation allows us to determine the offset distance the target workpiece has moved after being unloaded.

[0080] The greater the offset distance, the farther the distance between the actual center coordinates and the preset coordinates, and the more necessary it is to correct the deviation of the suction cup. When the mass of the workpiece accounts for a larger proportion, the position of the suction cup should be closer to the actual center coordinates of the workpiece (for deviation correction), so that the transfer of the workpiece will be more stable.

[0081] The aforementioned threshold can be obtained by statistically analyzing the average offset distance between the gripping position and the workpiece center when historical workpieces of different sizes are in an unstable state during the gripping process. The specific value can be determined based on the actual scenario. The unstable state refers to the historical workpiece being gripped and shaking during the transfer process without any board falling off.

[0082] Therefore, when the offset distance is greater than or equal to the threshold, a correction index needs to be calculated. That is, the correction index is negatively correlated with the degree of overlap and positively correlated with the quality percentage.

[0083] Specifically, corrective indicators for: ;

[0084] in, Let c be the degree of overlap of the c-th workpiece. Let c be the mass percentage of the c-th workpiece. It is an exponential function with the natural number e as the base. , Let be the length and width of the c-th workpiece, respectively. , These are the length and width of the suction cup, respectively.

[0085] When the suction cup is circular, the length and width of the suction cup mentioned above can be replaced with the diameter R. That is, the size of the workpiece and the suction cup is determined by comparing the length and width of the workpiece with the diameter.

[0086] The mass ratio is the ratio of the workpiece's weight to the maximum weight the suction cup can bear. The overlap ratio is the ratio of the suction cup's area to the area of ​​the workpiece within its projected area. When the overlap ratio equals 1, it means the suction cup can completely adhere to the workpiece; when the overlap ratio is less than 1, it means that part of the suction cup is not completely adhered to the workpiece, and the suction cup may need to be moved.

[0087] The weight of the workpiece can be estimated based on the unit weight per area. For example, if the area of ​​the workpiece is s square meters and the unit weight is g, then the weight of the workpiece is the product of s and g. It should be noted that the unit weight needs to be measured in advance.

[0088] The reason for the two scenarios for the aforementioned correction index is that some workpieces are relatively large, meaning the size of the workpiece is completely larger than the suction cup. In this case, if the position of the suction cup is slightly off, no correction is needed, and it will not have a significant impact on the transfer of the workpiece. However, if the deviation is large, it will cause the center of gravity to be too off-center, and the plate will fall off during the transfer process. On the other hand, some workpieces are relatively small, smaller than the length or width of the suction cup. In this case, the suction cup will leak air. If the air leakage area is too large, it will not be able to pick up the workpiece or will pick up the surrounding workpieces at an angle. Therefore, for the smaller workpieces, the suction cup should be in a more appropriate position. Hence, the correction index is set higher, meaning that correction is required whenever the suction cup deviates, so that it can transfer the workpiece more stably.

[0089] The threshold value is set at 0.6, but it can also be obtained by statistical analysis based on historical data of the suction cup during workpiece transfer.

[0090] Furthermore, before gripping the workpiece (including after the suction cup corrects its orientation), it is necessary to determine the degree of defect in the target workpiece area corresponding to the suction cup area. This only applies when the workpiece size is larger than the suction cup size. At the same time, a region image corresponding to the suction cup is acquired, and the region image is processed into grayscale to obtain a grayscale image; , Let be the length and width of the c-th workpiece, respectively. , The length and width of the suction cup are respectively

[0091] Acquire a region image corresponding to the suction cup area, and perform grayscale processing on the region image to obtain a grayscale image;

[0092] Obtain the difference image between the grayscale image and the standard image, perform thresholding on the difference image, and extract abnormal regions;

[0093] If the area of ​​the abnormal region exceeds the set value, the position of the suction cup will be readjusted.

[0094] The set value can be 5%. Of course, it can also be determined based on the actual situation. The above threshold processing can be the Otsu threshold processing method; the above readjustment of the suction cup position can be done by moving it clockwise or counterclockwise around the center coordinates to determine a suitable workpiece area.

[0095] The above-mentioned abnormal area area ratio is the ratio of the total number of pixels in the abnormal area to the total number of pixels in the area image.

[0096] Otsu's method is a global thresholding method based on grayscale histograms, also known as the maximum inter-class variance method. Its main idea is to divide the grayscale image into foreground and background parts, and determine the optimal grayscale threshold by maximizing the inter-class variance.

[0097] The aforementioned standard image is a defect-free workpiece image; it can be selected from multiple current workpiece images.

[0098] When the workpiece surface has dents or impurities that affect gripping, the working area of ​​the suction cup needs to be adjusted to exclude the abnormal area, thereby further improving the gripping success rate. And for items other than... For workpieces other than those that are small, such as those smaller than the length or width of the suction cup, the suction cup will leak air. If the leaking area is too large, the suction cup will not be able to pick up the board or will pick up the surrounding board at an angle. Therefore, for workpieces in this case, even if there are dents or impurities on the surface of the workpiece, the working area will not be adjusted.

[0099] In this embodiment, the specific process of using a suction cup to grasp the workpiece is as follows:

[0100] Step a1: Turn on the device, send the preset position corresponding to the preset coordinates of any workpiece to the robot, and move the suction cup to the preset position;

[0101] Step a2: Read the status of the suction cup. If the suction cup is in a stopped state, determine whether it has moved into place. If so, take a picture using the camera deployed on the suction cup and feed the captured workpiece image back to the processing module of the device to execute the above-mentioned suction cup correction method based on machine vision to complete the grasping of the target workpiece.

[0102] Step a3: After completing the workpiece gripping, send a command to transfer the workpiece to the position and release it, and determine whether the operation is complete. If it is complete, continue to send the robot the preset position corresponding to the preset coordinates of the next workpiece gripping by the suction cup, and so on, to achieve the gripping of all workpieces.

[0103] The present invention provides a suction cup correction method based on machine vision, which can judge the correction based on the deviation of the center coordinate when performing workpiece gripping. When there is almost no deviation, it can be gripped directly, making the gripping efficiency higher. When the deviation is large, the deviation of the suction cup position is judged by the degree of overlap and the proportion of workpiece mass, which improves the stability of subsequent workpiece transfer.

[0104] This invention also provides a suction cup alignment system based on machine vision. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the suction cup correction method based on machine vision according to the present invention.

[0105] 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 by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0106] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0107] While various embodiments of the invention have been shown and described in this specification, 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.

Claims

1. A suction cup correction method based on machine vision, characterized in that, include: Based on the preset coordinates of the target workpiece stored in the database, move the suction cup to the preset position; The image of the target workpiece is acquired, the center coordinates of the outline of the target workpiece are identified, and the center coordinates of the outline are converted into the true center coordinates of the target workpiece. The offset distance between the preset coordinates of the target workpiece and the actual center coordinates is obtained. If the offset distance is less than a threshold, the workpiece is directly grasped. If the offset distance is greater than or equal to the threshold, a correction index is calculated. If the correction index is greater than the set threshold, the preset position of the suction cup is corrected. Otherwise, the workpiece is directly grasped. The correction index represents the deviation of the suction cup. The suction cup is circular or rectangular in shape. When the suction cup is rectangular, the correction index... for: ； in, Let c be the degree of overlap of the c-th workpiece. Let c be the mass percentage of the c-th workpiece. It is an exponential function with the natural number e as the base. , Let be the length and width of the c-th workpiece, respectively. , These are the length and width of the suction cup, respectively. The degree of overlap is the ratio of the area of ​​the suction cup to the area of ​​the workpiece region in the projected area of ​​the c-th workpiece; the mass ratio is the ratio of the weight of the workpiece to the maximum weight that the suction cup can bear.

2. The suction cup correction method based on machine vision according to claim 1, characterized in that, The process of converting the center coordinates of the contour to the true center coordinates of the target workpiece involves converting pixel coordinates to the world coordinate system.

3. The suction cup correction method based on machine vision according to claim 2, characterized in that, The correction of the preset position of the suction cup includes: Calculate the offset angle between the preset coordinates and the actual center coordinates; The suction cup is rotated according to the offset angle.

4. The suction cup correction method based on machine vision according to claim 1, characterized in that, The true center coordinates are: ； in,( , ) represents the center coordinates of the profile of the c-th workpiece. , Let be the true center coordinates of the c-th workpiece after blanking, and K be the uniform scaling factor. Let be the relative rotation angle between the line connecting the center coordinates of the workpiece's contour and the theoretical center coordinates and the x-direction. , These are the x-direction translation vectors and y-direction translation vectors of the center coordinates of the contour before and after loading, relative to the theoretical center coordinates. , At least one of them is not 0; the theoretical center coordinates are the center coordinates of the workpiece image on the wooden board that has been laid out before cutting, and the wooden board is located on the xOy plane.

5. The suction cup correction method based on machine vision according to claim 4, characterized in that, The relative rotation angle for: ； in, , These are the x-direction translation vector and y-direction translation vector, respectively, between the center coordinates of the workpiece's contour in the image before and after loading and the theoretical center coordinates. It is an inverse cosine function.

6. The suction cup correction method based on machine vision according to claim 1, characterized in that, Before gripping the workpiece, the following steps are also included: like Then, the region image corresponding to the suction cup is acquired, and the region image is processed into grayscale to obtain a grayscale image; , Let be the length and width of the c-th workpiece, respectively. , These are the length and width of the suction cup, respectively. Obtain the difference image between the grayscale image and the standard image, perform thresholding on the difference image, and extract abnormal regions; If the area of ​​the abnormal region exceeds the set value, the position of the suction cup will be readjusted.

7. The suction cup correction method based on machine vision according to claim 1, characterized in that, The images were captured by a camera deployed on a suction cup.

8. The suction cup correction method based on machine vision according to claim 7, characterized in that, Also includes: The step of performing image enhancement processing on the image.

9. A suction cup alignment system based on machine vision, characterized in that, include: processor; A memory storing computer instructions for machine vision-based suction cup alignment, which, when executed by the processor, cause the system to perform a machine vision-based suction cup alignment method according to any one of claims 1-8.

Images