Stacked frame unstacking grabbing method, grabbing system and grabbing device based on 3D vision

Abstract

The invention discloses a stacking frame unstacking grabbing method, grabbing system and grabbing device based on 3D vision. The method comprises a method for grabbing a single-layer frame in the stacked frames, a method for unstacking and grabbing the stacked frames, an unstacking and grabbing system of the stacked frames and an unstacking and grabbing device of the stacked frames, wherein the unstacking and grabbing system of the stacked frames is arranged for achieving the unstacking and grabbing method, and the unstacking and grabbing device of the stacked frames is arranged for achieving the unstacking and grabbing method. According to the frame feeding device, different frames and different pieces of isolation paper can be accurately recognized and grabbed, the problem that materials fall off during grabbing is solved, multiple frames can be grabbed at the same time for unstacking operation, and the frame feeding efficiency is improved.

Description

[Technical field]

[0001] The invention belongs to the technical field of photovoltaic module production, and in particular relates to a destacking and grabbing method, a grabbing system and a grabbing device for stacking frames based on 3D vision. [Background technology]

[0002] During the production process of photovoltaic modules, it is necessary to assemble frames around the glass modules. The four frames are divided into two long frames and two end frames. The long frames and short frames are stacked together in multiple layers, and each layer has multiple frames. Since the front and back sides of the frames can be matched with each other, the frames are stacked in layers in the order of a layer of front-facing frame and a layer of back-facing frame. When the two layers of frames are stacked, the four sides are all flat. Therefore, the frames can be stacked in multiple layers and multiple frames in each layer to form a stack; in addition, in order to prevent friction between the frame layers, a frame is set between every two layers of frames. Isolation paper, in order to prevent both ends of the frame from being scratched, corner code isolation paper is set at both ends of the frame. Therefore, in the process of loading the frame, it is necessary to first grab the frame isolation paper and the corner code isolation paper, and then grab and move multiple front frames and back frames to the processing station. After taking away the same layer of frames many times, the next layer of frames is loaded. The frame isolation paper and the corner code isolation paper need to be taken away first. Therefore, correspondingly, our grasping system must be able to identify the frame isolation paper, corner code isolation paper, front frame and back frame. In order to improve the grasping efficiency, the grasping device is also required to grasp multiple frames at the same time.

[0003] In the prior art, for example, there is a depalletizing method, system, device, equipment and storage medium with Chinese patent publication number CN118343510A. In this solution, it is first determined whether the top layer of the cargo stack is a full layer; if the top layer is a full layer, the top layer posture is determined according to the package point cloud of each package, and a template posture matching the top layer posture is selected from the preset template postures as the first posture; then, based on the identified package point cloud, a grasping posture for overall grasping of all packages included in the top layer is determined; since each template posture corresponds to a preset grasping method, it can be The robot arm is controlled to grab all the packages included in the top layer from the stack to the specified position as a whole according to the determined grabbing posture and the grabbing method corresponding to the first posture. In this way, the robot arm can grab one layer of packages at a time for destacking. Although the scheme also applies 3D camera shooting, point cloud recognition, and package posture judgment, the above scheme takes away one layer of packages at a time. When the photovoltaic frames are stacked, the spacing between adjacent frames does not meet the requirements of the grabbing device. If the frames are taken away at one time, the positioning will be inaccurate during grabbing, resulting in the problem of frame falling off.

[0004] Therefore, it is necessary to provide a method, a system, and a device for unstacking and grasping the stacking frame based on 3D vision to solve the above technical problems.

Summary of the Invention

[0005] The main object of the present invention is to provide a method for grasping a single-layer frame in a stacking frame based on 3D vision, which can accurately complete the recognition and grasping of different frames, solve the problem of frame dropping during grasping, and at the same time can grasp multiple frames for unstacking operations, improving the efficiency of frame feeding.

[0006] The present invention realizes the above object through the following technical solutions: A method for grasping a single-layer frame of a stacking frame based on 3D vision, which includes the following steps:

[0007] S1. Denote the maximum number of frames that can be grasped by the frame grasping fixture at one time as m, and the number of frames in the single-layer frame as M. Divide all the frames in the single-layer frame into n frame units according to m, where n = [M / m] + 1;

[0008] S2. Use a 3D camera to take a picture of the set central area of the first frame unit, collect the point cloud data of m frames, and obtain the frame point cloud;

[0009] S3. Preprocess the frame point cloud, then match it with the set frame point cloud model to obtain the category information and pose information of the frame, and calculate the distance between every two adjacent frames among the 1 - m frames in the first frame unit; the category information of the frame includes the front frame and the back frame;

[0010] S4. Judge whether the distance between every two adjacent frames among the 1 - m frames meets the set distance range. If all meet, grasp the 1 - m frames according to the category information and pose information obtained in step S3, complete the grasping of the first frame unit, and execute step S7; otherwise, execute steps S5 - S6;

[0011] S5. Determine the frame combinations that do not meet the set distance range, and denote them as frame combination i = {frame A i , frame B i}, that is, the distance between the adjacent frame A i and frame B i does not meet the set distance range, where i = 1, 2, ……, p, 1 ≤ p < m - 1, and p is the number of frame combinations determined not to meet the set distance range;

[0012] S6. According to the category information and pose information obtained in step S3, grasp the 1st frame to frame A1 for the first time, then grasp frame B1 to frame A2 for the second time, and so on, until grasping frame B p to the mth frame to complete the grasping of the first frame unit;

[0013] S7, completing the capture of all subsequent frame units according to steps S2 to S6, and completing the capture of a single-layer frame.

[0014] Another main purpose of the present invention is to provide a method for destacking and grasping stacking frames based on 3D vision, which can accurately complete the identification and grasping of different frames and different isolation papers, and solve the problem of material falling during grasping.

[0015] The present invention achieves the above-mentioned purpose through the following technical solution: a method for destacking and grasping a stacking frame based on 3D vision, which comprises the following steps:

[0016] S100 and 3D cameras take photos of the top layer of the whole stack of frames to determine the overall height of the whole stack of frames;

[0017] S200, 3D camera takes pictures of the four corners of the whole stack frame, determines the center of the whole stack frame, and corrects the stack position of the whole stack frame;

[0018] S300, Grabbing of border isolation paper:

[0019] S301, the 3D camera moves to the top of the border isolation paper to take a picture of the border isolation paper, collects a point cloud of the border isolation paper, and obtains a point cloud of the border isolation paper;

[0020] S302, preprocessing the collected border isolation paper point cloud, matching the preprocessed border isolation paper point cloud with a preset border isolation paper point cloud model, and obtaining the position and posture information of the border isolation paper point cloud;

[0021] S303, according to the position information of the frame isolation paper obtained in step S302, select a corresponding grabbing fixture, grab the frame isolation paper and put it into the isolation paper material box;

[0022] S400, capturing a single-layer frame: completing the capture of a single-layer frame according to the single-layer frame capture method as described in claim 1;

[0023] S500, in the process of capturing the upper frame, further obtaining the height of the frame when taking a picture of the frame unit, adjusting the 3D camera to the optimal shooting height according to the height of the upper frame, and completing the calibration of the shooting height of the 3D camera;

[0024] S600, repeating steps S300 to S500 until all the frames of the stack are taken out, thus completing the unstacking of the stacked frames.

[0025] Another object of the present invention is to provide a stacking frame destacking and grabbing system based on 3D vision, which is used to complete the stacking frame destacking and grabbing method, which includes:

[0026] Centering photo correction module: the 3D camera takes photos of the four corners of the whole stack frame, determines the center of the whole stack frame, and corrects the stack position of the whole stack frame;

[0027] Point cloud acquisition module, used to obtain point cloud data and depth information of the frame and isolation paper;

[0028] The point cloud data processing module pre-processes the point cloud data and depth information of the point cloud acquisition module;

[0029] The point cloud matching module matches the preprocessed isolation paper point cloud with the preset isolation paper point cloud model to obtain the category information and posture information of the isolation paper. The category information of the isolation paper is frame isolation paper and corner code isolation paper; the preprocessed frame point cloud is matched with the preset frame point cloud model to obtain the type information and posture information of the frame, calculate the distance between each two adjacent frames in multiple frames, and determine whether the distance meets the requirements;

[0030] The posture conversion module uses the posture conversion matrix obtained by the hand-eye calibration model to convert the posture information of the border point cloud and the isolation paper point cloud into the grasping posture information of the robot gripper;

[0031] The grabbing device module receives the control instructions and posture data output by the posture conversion module, grabs the isolation paper and grabs the frame in sequence;

[0032] The photographing height correction module further obtains the height of the frame when photographing the frame unit during the upper frame capture process, adjusts the 3D camera to the optimal photographing height according to the height of the upper frame, and completes the correction of the photographing height of the 3D camera.

[0033] Another object of the present invention is to provide a destacking and grabbing device for stacking frames based on 3D vision, which is used to complete the above-mentioned destacking and grabbing method for stacking frames, and includes a mounting plate arranged at the movable end of the manipulator, a frame isolation paper grabbing clamp, a corner code isolation paper grabbing clamp and a frame grabbing clamp respectively arranged on the mounting plate.

[0034] Furthermore, the frame isolation paper grabbing fixture includes a first cylinder, a movable frame driven by the first cylinder to move up and down, and a plurality of first suction cups fixed at the lower end of the movable frame and used to absorb the frame isolation paper.

[0035] Furthermore, the corner code release paper grabbing fixture includes a second cylinder and a pair of clamping claws driven by the second cylinder to open or clamp and used to clamp the corner code release paper.

[0036] Furthermore, the frame clamping fixture includes a plurality of air circuit blocks installed on the mounting plate and arranged at equal intervals along a first direction, a plurality of second suction cups fixedly arranged along a second direction at the lower end of the air circuit block, and a motor driving the plurality of air circuit blocks to simultaneously increase or simultaneously decrease the spacing along the first direction. The first direction is perpendicular to the second direction. When the frame clamping fixture absorbs the frame, the first direction is the width direction of the frame, and the second direction is the length direction of the frame.

[0037] Compared with the prior art, the stacking frame destacking grasping method, grasping system and grasping device of the present invention based on 3D vision have the following beneficial effects:

[0038] (1) Before grasping, the type and position of the frame and the isolation paper can be identified according to the point cloud data collected by the 3D camera, and the spacing information between each two adjacent frames among multiple frames can be calculated, and it can be determined whether the spacing between the two adjacent frames meets the requirements; the frames with spacing meeting the requirements are taken away in turn, and the grasping fixture can take away multiple frames each time, which can improve the efficiency of frame destacking;

[0039] (2) The grabbing system is provided with a photographing height correction module and a centering photographing correction module. The photographing height correction module can output the frame stacking height information according to the frame point cloud collected last time, adjust the 3D camera to the optimal photographing height, and correct the photographing height of the whole stack of frames to ensure that the photographing field of view height of the 3D camera is within the range; in the centering photographing correction module, the 3D camera photographs the four corners of the whole stack of frames, determines the center of the whole stack of frames, and corrects the stacking position of the whole stack of frames, which can ensure that the frames can enter the photographing field of view, and can photograph the central area of ​​the frames to ensure accurate photographing positioning;

[0040] (3) The grasping device integrates the frame isolation paper grasping fixture, the corner code isolation paper grasping fixture and the frame grasping fixture into one robot, which can save robots and can also quickly select different grasping fixtures according to grasping needs, thereby improving the grasping rhythm.

[0041] (4) When matching the collected point cloud with the pre-built point cloud model, the point cloud is first roughly matched based on the fast point feature histogram (FPFH), and then the iterative closest point algorithm (ICP) is used to fine-tune the pose result. The coarse matching can provide a reliable initial position for the precise matching of the iterative closest point algorithm (ICP), thereby accelerating the registration speed of the iterative closest point algorithm (ICP), achieving precise positioning of the border and isolation paper, and ensuring the accuracy of the grasping position, solving the problem of frame loss due to low grasping positioning accuracy.

Brief Description of the Drawings

[0042] Figure 1It is a schematic diagram of the three-dimensional structure of a grabbing device according to an embodiment of the present invention;

[0043] Figure 2 It is a schematic diagram of the three-dimensional structure of a grabbing device according to an embodiment of the present invention;

[0044] Figure 3 It is a schematic diagram of the three-dimensional structure of the frame grabbing fixture according to an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of the three-dimensional structure of the release paper grabbing fixture according to an embodiment of the present invention;

[0046] Figure 5 The steps of the method for destacking and grabbing the stacking frame according to the embodiment of the present invention;

[0047] The numbers in the figure represent:

[0048] 3-grabbing device, 31-mounting plate, 32-frame grabbing fixture, 321-air circuit block, 322-second suction cup, 323-motor, 33-frame isolation paper grabbing fixture, 331-first cylinder, 332-moving frame, 333-first suction cup, 34-corner code isolation paper grabbing fixture, 341-second cylinder, 342-clamping claw, 35-3D camera, 36-laser rangefinder. [Specific implementation method]

[0049] In the process of destacking the frames, it is necessary to first grab the frame isolation paper and the corner code isolation paper, and then grab and move multiple front frames and back frames to the processing station. After taking away the frames of the same layer multiple times, before loading the next layer of frames, it is necessary to take away the frame isolation paper and the corner code isolation paper. Therefore, in the process of loading the frames, it is necessary to take away the isolation paper on the frames before moving the frames. Therefore, it is necessary to grab the isolation paper and the frames during the process of the frames being loaded. Therefore, our grabbing device must be able to grab the frame isolation paper, the corner code isolation paper and the frames at the same time. Correspondingly, Figure 1-Figure 4 As shown, the 3D vision-based stacking frame destacking grasping device 3 includes a manipulator, a mounting plate 31 arranged at the movable end of the manipulator, a frame isolation paper grasping clamp 33, a corner code isolation paper grasping clamp 34 and a frame grasping clamp 32 respectively arranged on the mounting plate 31.

[0050] The frame isolation paper grabbing fixture 33 includes a first cylinder 331, a moving frame 332 driven by the first cylinder 331 to move up and down, and a plurality of first suction cups 333 fixed at the lower end of the moving frame 332 and used to absorb the frame isolation paper.

[0051] The corner code release paper grabbing fixture 34 includes a second cylinder 341 and a pair of clamping claws 342 driven by the second cylinder 341 to open or clamp and to clamp the corner code release paper.

[0052] The frame clamping fixture 32 includes a plurality of air circuit blocks 321 arranged at equal intervals along a first direction and mounted on a mounting plate 31, a plurality of second suction cups 322 fixedly arranged along a second direction at the lower end of each air circuit block 321, and a motor 323 driving the plurality of air circuit blocks 321 to simultaneously increase or simultaneously decrease the spacing along the first direction. The first direction and the second direction are perpendicular to each other. When the frame clamping fixture 32 absorbs the frame, the first direction is the width direction of the frame, and the second direction is the length direction of the frame. In order to ensure the stability of the grasping, the second suction cup 322 is adsorbed on the center line of the short side of the frame. If the adsorption position deviates from the center line of the short side, the frame may fall off.

[0053] The motor 323 drives several gas circuit blocks 321 to move along the first direction at the same time, and can simultaneously reduce the spacing between several gas circuit blocks 321 or simultaneously increase the spacing between several gas circuit blocks 321 to adjust the spacing between several second suction cups 322 in the first direction, so as to be suitable for multiple synchronous adsorption of frames of different widths, which can not only improve the efficiency of transportation, but also have higher versatility and good flexibility. In this embodiment, ten gas circuit blocks 321 are arranged at equal intervals along the first direction, and ten frames can be adsorbed at the same time to improve the efficiency of adsorption. In other embodiments, the number of gas circuit blocks 321 can be adjusted according to actual conditions and is not limited here. Several second suction cups 322 fixedly arranged along the second direction at the lower end of the gas circuit block 321 are adsorbed on the same frame at the same time, which can improve the accuracy and reliability of adsorption.

[0054] In order to ensure the accuracy and reliability of the frame absorption, two sets of frame clamping fixtures 32 are provided and are respectively arranged at the two ends of the mounting plate 31, and the two sets of frame clamping fixtures 32 are respectively adsorbed on the left and right ends of the frame. A 3D camera 35 and a laser rangefinder 36 are arranged on the mounting plate 31, and the 3D camera 35 and the laser rangefinder 36 are arranged in the middle of the two sets of frame clamping fixtures 32. In this embodiment, the 3D camera 35 is a laser galvanometer camera, and the laser rangefinder 36 measures the height and plane spacing.

[0055] When destacking the frames, each layer is destacking first, until the bottom layer is taken away, the whole stack of frames is considered destacking. Below, we first explain the destacking method of a single-layer frame. The single-layer frame here refers to a front frame and a back frame. First take away a layer of the front frame, then take away a layer of the back frame, and the grabbing method of the front frame and the back frame is the same. The single-layer frame grabbing method in the stacking frame based on 3D vision, the steps of grabbing a single-layer frame with a grabbing fixture are as follows:

[0056] S1. The maximum number of frames that the frame grabbing fixture can grab at one time is m, the number of frames in a single-layer frame is M, and all frames in the single-layer frame are divided into n frame units according to m, n = [M / n] + 1; wherein n is rounded upwards to an integer; and M frames are arranged in sequence;

[0057] S2, using a 3D camera to take a picture of the set central area of ​​the first border unit, collect point cloud data of m borders, and obtain a border point cloud;

[0058] The frame grabbing fixture can grab m frames at most. Since the spacing between frames may increase when each layer of frames is arranged, in order to ensure that as many m frames as possible are grabbed, the 3D camera's field of view needs to be increased when taking pictures. The field of view needs to be larger than the field of view where the m frames are located. Therefore, in order to facilitate calculation and understanding, in this embodiment, the 3D camera's field of view is calculated by the field of view occupied by k frames, where m < k. At the same time, it can be ensured that when all frames are offset to the maximum, it can still meet the requirement of grabbing m frames at most. In other embodiments, calculation can be performed in other ways, as long as the 3D camera's field of view can ensure that when all frames are offset to the maximum, it can still collect the point cloud of m frames.

[0059] S3, preprocessing the border point cloud, and then matching it with the set border point cloud model, obtaining the category information and the pose information of the border, and calculating the distance between each two adjacent borders in the 1st to mth borders of the first border unit; the category information of the border includes the front border and the back border;

[0060] Specifically, in step S3, the types of frames include front frames and back frames. Since the front and back sides of the frames can be matched with each other, the frames are stacked in layers in the order of a front frame facing up and a back frame facing up. When the two layers of frames are stacked, the four sides are all flat. Therefore, the frames can be stacked in multiple layers and multiple frames in each layer. The frames include long frames and short frames. Each glass component requires two long frames and a short frame to be assembled together. Both the long frames and the short frames are distinguished by the front frame and the back frame. The long frames and the short frames have the same stacking method and the same grabbing method.

[0061] When matching the collected border point cloud with the border point cloud model, the collected border point cloud needs to be preprocessed first. The preprocessing includes denoising, filtering, downsampling, etc., to improve the efficiency and accuracy of subsequent processing; after the preprocessing is completed, the point cloud is roughly matched based on the fast point feature histogram (FPFH) to obtain the initial pose information, and then the iterative closest point algorithm (ICP) is used to fine-tune the initial pose information. After fine-tuning, the category information of the border and the pose information of the border points are output. The rough matching can provide a reliable initial position for the precise matching of the iterative closest point algorithm (ICP), thereby accelerating the registration speed of the iterative closest point algorithm (ICP) and realizing the rapid positioning of the border. After the iterative closest point algorithm (ICP) matching, the specific position of the border in the point cloud data can be obtained, including the center point coordinates and posture (such as rotation angle) of the object.

[0062] Before grabbing the border, you need to build a border point cloud model. The construction method of the border point cloud model is:

[0063] S31, scanning a single border to obtain a border point cloud set;

[0064] S32, preprocessing the border point cloud set, the preprocessing including denoising, filtering, downsampling, etc.;

[0065] S33, identifying and saving the type of the border, and marking the grab points in the border point cloud set to obtain a border point cloud model, and saving the border point cloud model.

[0066] Since the air path blocks 321 on the grabbing fixture are arranged at equal intervals in the first direction when the grabbing fixture is grabbing, the second suction cups 322 are also arranged at equal intervals in the first direction. Therefore, when sucking the frame, the spacing between the center lines of the short sides of the two adjacent frames being sucked is consistent and consistent with the spacing of the second suction cups 322 in the first direction. If the spacing between a frame and the adjacent frame becomes smaller or larger and does not meet the spacing requirements and an abnormal position occurs, it will cause several second suction cups 322 to fail to be adsorbed on the center line of the short side of the frame, resulting in the adsorption position deviation and the frame falling off, affecting the frame grabbing operation. Therefore, it is necessary to obtain the spacing information between each two adjacent frames in the frame unit to confirm whether the spacing between each two adjacent frames meets the requirements, and the grabbing fixture adopts different grabbing methods according to the spacing information.

[0067] S4, judging whether the spacing between each two adjacent frames in 1 to m frames meets the set spacing range, if all meet, grabbing 1 to m frames according to the category information and posture information obtained in step S3, completing the grabbing of the first frame unit, and executing step S7; otherwise, executing steps S5 to S6;

[0068] S5. Determine the border combinations that do not meet the set spacing range, and denote them as border combination i = {border A i , border B i}, that is, the spacing between adjacent border A i and border B i does not meet the set spacing range, i = 1, 2, ……, p, 1 ≤ p < m - 1, and p is the number of border combinations determined not to meet the set spacing range;

[0069] S6. According to the category information and pose information obtained in step S3, grab the first to border A1 for the first time, then grab border B1 to border A2 for the second time, and so on, until grabbing border B p to the mth border, and complete the grabbing of the first border unit;

[0070] S7. Complete the grabbing of all subsequent border units according to steps S2 - S6 to complete the grabbing of a single layer of borders.

[0071] Since there is separator paper between the upper and lower single - layer borders, when grabbing the entire stack of borders, the border separator paper needs to be removed first before grabbing a single layer of borders. Therefore, the method for depalletizing photovoltaic borders based on 3D vision, as Figure 5 shown, has the following steps:

[0072] S100. The 3D camera takes a photo of the top layer of the entire stack of borders to determine the overall height of the entire stack of borders; since the number of layers of the incoming entire stack of borders is different, resulting in inconsistent overall heights, it is necessary to first use the 3D camera to take a photo to confirm the overall height of the entire stack of borders, so as to determine the optimal photo - taking height, ensure that the photo - taking vision height of the 3D camera is within the range, and make the photo - taking clear for subsequent point cloud acquisition;

[0073] S200. The 3D camera takes photos of the four corners of the entire stack of borders to determine the center of the entire stack of borders and correct the stack position; when the entire stack of borders is placed in the feeding area, there may be a position offset. If the 3D camera directly takes a photo to collect the border point cloud, it will cause an offset during photo - taking, which may result in some borders not entering the photo - taking vision or being unable to take a photo of the central area of the border. By using the 3D camera to take photos of the four corners of the entire stack of borders to determine the center of the entire stack of borders, by determining the center position of the entire stack of borders, it is possible to move above the corresponding border for photo - taking, which can ensure that the border can enter the photo - taking vision and can take a photo of the central area of the border, ensuring accurate photo - taking positioning;

[0074] S300. Grabbing the border separator paper, which includes the following steps:

[0075] S301. The 3D camera moves above the border separator paper to take a photo of the separator paper and collect the point cloud of the border separator paper to obtain the point cloud of the border separator paper;

[0076] S302, preprocessing the collected border isolation paper point cloud, matching the preprocessed border isolation paper point cloud with a preset border isolation paper point cloud model, and obtaining the position and posture information of the border isolation paper point cloud;

[0077] Specifically, in step S302, the preprocessing of the border isolation paper point cloud includes denoising, filtering, downsampling, etc.; before the border isolation paper is captured, it is necessary to first construct a border isolation paper point cloud model of the border isolation paper, and the construction method of the border isolation paper point cloud model is:

[0078] S3021, scanning the border isolation paper to obtain a point cloud set of the border isolation paper;

[0079] S3022, preprocessing the border isolation paper point cloud set, the preprocessing including denoising, filtering, downsampling, etc.;

[0080] S3023, identifying and saving the type of the border isolation paper, and marking the grab points in the border isolation paper point cloud set to obtain a border isolation paper point cloud model, and saving the border isolation paper point cloud model.

[0081] S303, according to the outputted position information of the frame isolation paper, select a corresponding grabbing fixture, grab the frame isolation paper and put it into the isolation paper material box;

[0082] S400, grabbing a single-layer frame, and completing the grabbing of the single-layer frame according to the grabbing method of the single-layer frame in the above steps S1 to S7; the single-layer frame referred to here refers to a layer of front frame and a layer of back frame, first take away a layer of front frame, and then take away a layer of back frame, and the grabbing method of the front frame and the back frame is the same, so it is not repeated.

[0083] S500, in the process of capturing the upper frame, further obtaining the height of the frame when taking a picture of the frame unit, adjusting the 3D camera to the optimal shooting height according to the height of the upper frame, and completing the calibration of the shooting height of the 3D camera;

[0084] Specifically, in step S500, after removing a single layer of the frame, the height of the entire stack of materials will be reduced. Therefore, during the acquisition of the isolation paper point cloud and the frame point cloud, the robot will be guided to descend according to the height information returned by the point cloud, thereby ensuring that the 3D camera's shooting height is within the optimal range. Therefore, during the process of grabbing the frame, the stack position of the frame will be lowered, and the 3D camera's shooting height will be reduced accordingly.

[0085] S600, repeating steps S300 to S500 until all the frames of the stack are taken out, thus completing the unstacking of the stacked frames.

[0086] Since there are corner codes at both ends of the short frame, corner code separation paper is also set at both ends of the short frame. Therefore, it is necessary to add a step of grabbing the corner code separation paper before grabbing the short frame. Therefore, before the above step S400 grabs the single-layer frame, there is also a step S700 to grab the corner code separation paper. The step of grabbing the corner code separation paper in S700 is as follows:

[0087] S701, the 3D camera moves to the top of the corner code isolation paper to take a picture of the corner code isolation paper, collects a point cloud of the corner code isolation paper, and obtains a point cloud of the corner code isolation paper;

[0088] S702, preprocessing the collected angle code isolation paper point cloud, matching the preprocessed angle code isolation paper point cloud with a preset angle code isolation paper point cloud model, and obtaining the position and posture information of the angle code isolation paper point cloud;

[0089] S703. According to the position information of the corner code isolation paper obtained in step S702, a corresponding grabbing fixture is selected to grab the corner code isolation paper and put it into an isolation paper material box.

[0090] In step S702, the method for constructing the corner code isolation paper point cloud model is consistent with the method for constructing the border isolation paper point cloud model, which will not be described again here.

[0091] In order to complete the above-mentioned stacking frame depalletizing and grasping method based on 3D vision, a stacking frame depalletizing and grasping system based on 3D vision is also provided, which includes:

[0092] Centering photo correction module: the 3D camera takes photos of the four corners of the whole stack frame, determines the center of the whole stack frame, and corrects the stack position of the whole stack frame;

[0093] Point cloud acquisition module, used to obtain point cloud data and depth information of the frame and isolation paper;

[0094] The point cloud data processing module pre-processes the point cloud data and depth information of the point cloud acquisition module;

[0095] The point cloud matching module matches the preprocessed isolation paper point cloud with the preset isolation paper point cloud model to obtain the category information and posture information of the isolation paper. The category information of the isolation paper is frame isolation paper and corner code isolation paper; the preprocessed frame point cloud is matched with the preset frame point cloud model to obtain the type information and posture information of the frame, calculate the distance between each two adjacent frames in multiple frames, and determine whether the distance meets the requirements;

[0096] The posture conversion module converts the posture information of the frame point cloud and the isolation paper point cloud into the grasping posture information of the robot gripper by using the posture conversion matrix obtained by the hand-eye calibration model; wherein, the posture conversion matrix of the isolation paper or the frame is output by the hand-eye calibration model, and the posture of the isolation paper or the frame in the 3D camera can be converted into the posture in the grasping device module by the posture conversion matrix; the acquisition of the posture conversion matrix requires the hand-eye calibration of the 3D camera, and the eye-in-hand calibration method is adopted, which is a prior art and will not be described here;

[0097] The grabbing device module receives the control instructions and posture data output by the posture conversion module, executes the grabbing action, and grabs the isolation paper and the frame;

[0098] The photographing height correction module further obtains the height of the frame when photographing the frame unit during the upper frame capture process, adjusts the 3D camera to the optimal photographing height according to the height of the upper frame, and completes the correction of the photographing height of the 3D camera.

[0099] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for capturing a single-layer frame in a stacking frame based on 3D vision, characterized in that: It includes the following steps: S1. The maximum number of frames that the frame grabbing fixture can grab at one time is m, the number of frames in a single-layer frame is M, and all frames in the single-layer frame are divided into n frame units according to m, where n = [M / m] + 1; S2, using a 3D camera to take a picture of the set central area of ​​the first border unit, collect point cloud data of m borders, and obtain a border point cloud; S3, preprocessing the border point cloud, and then matching it with the set border point cloud model, obtaining the category information of the border and the position information of the border, and calculating the distance between each two adjacent borders in the first border unit 1 to m borders; the category information of the border includes the front border and the back border; S4, judging whether the spacing between each two adjacent frames in 1 to m frames meets the set spacing range, if all meet, grabbing 1 to m frames according to the category information and posture information obtained in step S3, completing the grabbing of the first frame unit, and executing step S7; otherwise, executing steps S5 to S6; S5. Determine the border combinations that do not meet the set spacing range, and denote them as border combination i = {border A i , border B i}, that is, the spacing between adjacent border A i and border B i does not meet the set spacing range, i = 1, 2, ……, p, 1 ≤ p < m - 1, where p is the number of border combinations determined not to meet the set spacing range; S6, according to the category information and posture information obtained in step S3, the first frame A1 is captured, and then the frames B1 to A2 are captured for the second time, and so on, until the frame B is captured. p ~The mth border, complete the capture of the first border unit; S7, completing the capture of all subsequent frame units according to steps S2 to S6, and completing the capture of a single-layer frame.

2. A method for destacking and grabbing a stacking frame based on 3D vision, characterized in that: It includes the following steps: S100 and 3D cameras take photos of the top layer of the whole stack of frames to determine the overall height of the whole stack of frames; S200, 3D camera takes pictures of the four corners of the whole stack frame, determines the center of the whole stack frame, and corrects the stack position of the whole stack frame; S300, Grabbing of border isolation paper: S301, the 3D camera moves to the top of the border isolation paper to take a picture of the border isolation paper, collects a point cloud of the border isolation paper, and obtains a point cloud of the border isolation paper; S302, preprocessing the collected border isolation paper point cloud, matching the preprocessed border isolation paper point cloud with a preset border isolation paper point cloud model, and obtaining the position and posture information of the border isolation paper point cloud; S303, according to the position information of the frame isolation paper obtained in step S302, select a corresponding grabbing fixture, grab the frame isolation paper and put it into the isolation paper material box; S400, capturing a single-layer frame: completing the capture of a single-layer frame according to the single-layer frame capture method as described in claim 1; S500, in the process of capturing the upper frame, further obtaining the height of the frame when taking a picture of the frame unit, adjusting the 3D camera to the optimal shooting height according to the height of the upper frame, and completing the calibration of the shooting height of the 3D camera; S600, repeating steps S300 to S500 until all the frames of the stack are taken out, thus completing the unstacking of the stacked frames.

3. The method for destacking and grasping a stacking frame based on 3D vision as claimed in claim 2, characterized in that: Before the above step S400 of grabbing the single-layer frame, there is also step S700 of grabbing the corner code isolation paper. The steps of grabbing the corner code isolation paper in step S700 are as follows: S701, the 3D camera moves to the top of the corner code isolation paper to take a picture of the corner code isolation paper, collects a point cloud of the corner code isolation paper, and obtains a point cloud of the corner code isolation paper; S702, preprocessing the collected angle code isolation paper point cloud, matching the preprocessed angle code isolation paper point cloud with a preset angle code isolation paper point cloud model, and obtaining the position and posture information of the angle code isolation paper point cloud; S703. According to the position information of the corner code isolation paper obtained in step S702, a corresponding grabbing fixture is selected to grab the corner code isolation paper and put it into an isolation paper material box.

4. The stacking frame destacking and grabbing system based on 3D vision is characterized by The method for destacking and grabbing the stacking frame as claimed in claim 2 comprises: Centering photo correction module: the 3D camera takes photos of the four corners of the whole stack frame, determines the center of the whole stack frame, and corrects the stack position of the whole stack frame; Point cloud acquisition module, used to obtain point cloud data and depth information of the frame and isolation paper; The point cloud data processing module pre-processes the point cloud data and depth information of the point cloud acquisition module; The point cloud matching module matches the preprocessed isolation paper point cloud with the preset isolation paper point cloud model to obtain the category information and posture information of the isolation paper. The category information of the isolation paper is frame isolation paper and corner code isolation paper; the preprocessed frame point cloud is matched with the preset frame point cloud model to obtain the type information and posture information of the frame, calculate the distance between each two adjacent frames in multiple frames, and determine whether the distance meets the requirements; The posture conversion module uses the posture conversion matrix obtained by the hand-eye calibration model to convert the posture information of the border point cloud and the isolation paper point cloud into the grasping posture information of the robot gripper; The grabbing device module receives the control instructions and posture data output by the posture conversion module, grabs the isolation paper and the frame in turn; The photographing height correction module further obtains the height of the frame when photographing the frame unit during the upper frame capture process, adjusts the 3D camera to the optimal photographing height according to the height of the upper frame, and completes the correction of the photographing height of the 3D camera.

5. A 3D vision-based stacking frame destacking and grabbing device, characterized in that The method for destacking and grabbing the stacking frame as described in claim 2 includes a mounting plate arranged at the movable end of the manipulator, a frame isolation paper grabbing clamp, a corner code isolation paper grabbing clamp and a frame grabbing clamp respectively arranged on the mounting plate.

6. The photovoltaic frame destacking and grabbing device based on 3D vision according to claim 5, characterized in that: The frame isolation paper grabbing fixture comprises a first cylinder, a moving frame driven by the first cylinder to move up and down, and a plurality of first suction cups fixed at the lower end of the moving frame and used to absorb the frame isolation paper.

7. The photovoltaic frame destacking and grabbing device based on 3D vision according to claim 5, characterized in that: The corner code isolation paper grabbing fixture comprises a second cylinder and a pair of clamping claws driven by the second cylinder to open or clamp and used to clamp the corner code isolation paper.

8. The photovoltaic frame destacking and grabbing device based on 3D vision according to claim 5, characterized in that: The frame clamping fixture includes a plurality of air circuit blocks installed on the mounting plate and arranged at equal intervals along a first direction, a plurality of second suction cups fixedly arranged along a second direction at the lower end of the air circuit blocks, and a motor driving the plurality of air circuit blocks to simultaneously increase or simultaneously decrease the spacing along the first direction. The first direction is perpendicular to the second direction. When the frame clamping fixture absorbs the frame, the first direction is the width direction of the frame, and the second direction is the length direction of the frame.

Images