Industrial robot optical navigation anti-shielding tracking system, method and device and medium

Abstract

The invention discloses an industrial robot optical navigation anti-shielding tracking system, method and device and a medium, the system comprises a sensing layer, a calculation layer and an execution layer, the sensing layer comprises a mixed vision network composed of two active pan-tilt cameras and four fixed cameras and is used for obtaining a global workpiece image of a workpiece, and the calculation layer comprises a first camera and a second camera. Marking points are arranged on the workpiece; the calculation layer comprises a pose calculation module, a motion prediction module and a fusion decision module, and is used for receiving the image data of the sensing layer and processing the image data to obtain pose data; and the execution layer comprises an industrial robot and is used for generating a control instruction based on the pose data so as to transmit the control instruction to a control end of the industrial robot. Through the dynamic viewpoint optimization, the normal direction constraint modeling and the hierarchical motion prediction algorithm, the tracking precision and the real-time performance of the industrial robot under the complex working condition are remarkably improved, meanwhile, the cost and the power consumption are reduced, and the method has wide application prospects.

Description

Technical Field

[0001] The present invention relates to the field of industrial robot visual navigation, and more specifically, to an industrial robot optical navigation anti-occlusion tracking system, method, device and medium. Background Art

[0002] In the field of intelligent manufacturing, industrial robots' real-time and accurate tracking of workpieces is a key technology for achieving automated assembly, welding and other processes. Traditional optical navigation systems have three major technical bottlenecks: 1. Field of view limitations: Commercial optical tracking systems (such as NDI Polaris) typically have a measurement volume (MV) of less than 2m³. Tracking will be interrupted when the workpiece moves beyond the MV. Research shows that in automotive welding scenarios, the tracking failure rate due to workpiece movement is as high as 32%.

[0003] 2. Dynamic occlusion: Temporary occlusion caused by robots, lifting equipment, and other factors is common in industrial sites. Existing multi-camera solutions (such as CN110682245A) expand the field of view but fail to resolve multi-camera data conflicts. Positioning errors increase by 3-5 times under 50% occlusion.

[0004] 3. Posture sensitivity: When the angle between the workpiece normal and the camera optical axis exceeds 60°, the success rate of marker recognition drops below 65%. Existing techniques (such as Wang et al. 2016) only consider position constraints and ignore posture constraints, resulting in tracking failure during rapid rotation.

[0005] Existing solutions have obvious shortcomings. Among them, the fixed multi-camera layout proposed in patent CN112987732A cannot adapt to dynamic working conditions; patent CN111693018B adopts an active navigation solution with a camera mounted on a robotic arm, but does not establish a normal direction constraint model, and the error reaches ±2.3mm under the workpiece flipping condition; although the literature (Han et al. 2023) introduces relative velocity coordinates, the calculation delay reaches 120ms, which cannot meet real-time control requirements. Summary of the Invention

[0006] The purpose of the present invention is to provide an industrial robot optical navigation anti-occlusion tracking system, method, device and medium, which are particularly suitable for real-time high-precision tracking of workpieces under dynamic occlusion and complex motion conditions.

[0007] A first aspect of the present invention provides an industrial robot optical navigation anti-occlusion tracking system, comprising: Perception layer, computing layer and execution layer, among which, The perception layer includes a hybrid vision network composed of two active pan-tilt cameras and four fixed cameras, for acquiring a global workpiece image of a workpiece, wherein the workpiece is provided with a marker point; The calculation layer includes a pose solving module, a motion prediction module and a fusion decision module, for receiving and processing image data of the perception layer to obtain pose data; The execution layer includes an industrial robot, for generating a control instruction based on the pose data to be transmitted to a control end of the industrial robot.

[0008] In the scheme, the pose solving module is used to identify the marker point in the workpiece image; the motion prediction module is used to correct the workpiece motion position; and the fusion decision module is used to detect and eliminate abnormal poses by Mahalanobis distance, and to calculate the pose data by fusing multi-camera data.

[0009] The second aspect of the application provides an industrial robot optical navigation anti-occlusion tracking method, applied to any one of the industrial robot optical navigation anti-occlusion tracking systems, including the following steps: Collecting multi-camera data for preprocessing to obtain a target image, wherein the multi-camera includes two active pan-tilt cameras and four fixed cameras; Performing pose calculation, motion prediction and adaptive weighted fusion on the target image to obtain pose data; Based on the pose data, a control instruction is generated for the target angle of each joint of the robot, so as to control the joint action of the robot based on the control instruction.

[0010] In the scheme, the target image is subjected to pose calculation and motion prediction, specifically including: The target image is extracted, including a de-distorted image and 2D pixel coordinates; Based on the 2D pixel coordinates and the preset 3D model coordinates, the 6D pose of the workpiece is calculated; Based on the 6D pose of the workpiece, the motion prediction is performed to correct the workpiece motion position to obtain a predicted pose, wherein the Newton iteration correction is specifically performed.

[0011] In the scheme, adaptive weighted fusion is performed to obtain the pose data, specifically including: Based on the 6D pose and the predicted pose, the visibility score of each camera to the workpiece is calculated; Based on the visibility score, the view angle of the multi-camera is adjusted, and the pose data is calculated by adaptive weighted fusion based on the confidence of different cameras.

[0012] A third aspect of the present invention further provides an industrial robot optical navigation anti-occlusion tracking device, comprising a memory and a processor, wherein the memory includes an industrial robot optical navigation anti-occlusion tracking method program, and when the industrial robot optical navigation anti-occlusion tracking method program is executed by the processor, the following steps are implemented: Collecting multi-camera data and preprocessing it to obtain the target image, where the multi-camera includes two active pan-tilt cameras and four fixed cameras; Performing pose calculation, motion prediction, and adaptive weighted fusion on the target image to obtain pose data; The target angle of each joint of the robot is calculated based on the posture data to generate a control instruction, thereby controlling the robot joint movement based on the control instruction.

[0013] In this solution, pose calculation and motion prediction are performed on the target image, specifically including: Extracting the target image, wherein the target image includes a dedistorted image and 2D pixel coordinates; Calculate the workpiece's 6D pose based on 2D pixel coordinates and preset 3D model coordinates; Motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose, wherein the correction is specifically performed iteratively through the Newton method.

[0014] In this solution, the pose data is obtained by adaptive weighted fusion, which specifically includes: Calculating a visibility score of the workpiece for each camera based on the 6D pose and the predicted pose; The multi-camera viewing angles are adjusted based on the visibility scores, and the pose data are obtained by performing adaptive weighted fusion calculation based on the confidence levels of different cameras.

[0015] The fourth aspect of the present invention provides a computer-readable storage medium, which includes an industrial robot optical navigation anti-occlusion tracking method program for a machine. When the industrial robot optical navigation anti-occlusion tracking method program is executed by a processor, the steps of an industrial robot optical navigation anti-occlusion tracking method as described in any one of the above items are implemented.

[0016] The present invention discloses an industrial robot optical navigation anti-occlusion tracking system, method, device and medium. Through dynamic viewpoint optimization, normal direction constraint modeling and hierarchical motion prediction algorithm, the system significantly improves the tracking accuracy and real-time performance of industrial robots under complex working conditions, while reducing costs and power consumption, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A block diagram of an optical navigation and anti-occlusion tracking system for an industrial robot according to the present invention is shown; Figure 2 A flow chart of an industrial robot optical navigation anti-occlusion tracking method according to the present invention is shown; Figure 3 A schematic diagram of a normal direction constraint set model of an industrial robot optical navigation anti-occlusion tracking method according to the present invention is shown; Figure 4 A block diagram of an optical navigation and anti-occlusion tracking device for an industrial robot according to the present invention is shown. DETAILED DESCRIPTION

[0018] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0019] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0020] Figure 1 A block diagram of an industrial robot optical navigation anti-occlusion tracking system of the present application is shown.

[0021] like Figure 1 As shown, the present application discloses an industrial robot optical navigation anti-occlusion tracking system, comprising: Perception layer, computing layer and execution layer, among which, The perception layer includes a hybrid vision network composed of two active pan-tilt cameras and four fixed cameras, and is used to obtain a global workpiece image of the workpiece, wherein the workpiece is provided with marking points; The computing layer includes a posture solving module, a motion prediction module and a fusion decision module, which are used to receive the image data of the perception layer and process it to obtain posture data; The execution layer includes an industrial robot, which is used to generate control instructions based on the posture data and transmit them to the control end of the industrial robot.

[0022] It should be noted that, in this embodiment, Figure 1 As shown, the perception layer includes a pan-tilt camera and a fixed camera, specifically a hybrid vision network consisting of two active pan-tilt cameras and four fixed cameras, which is used to obtain a global workpiece image of the workpiece, wherein the workpiece is provided with marking points. In one embodiment of the invention, the frame rate of the active pan-tilt camera is 200 Hz, the focal length is 12 mm, the frame rate of the fixed camera is 50 Hz, and the coverage space of the hybrid vision network includes 8 m³.

[0023] Furthermore, in this embodiment, the computing layer includes a posture solving module, a motion prediction module and a fusion decision module, which are used to receive the image data of the perception layer and process it to obtain posture data, wherein the posture solving module is used to identify the marking points in the workpiece image, the motion prediction module is used to correct the motion position of the workpiece, and the fusion decision module is used to eliminate abnormal postures through Mahalanobis distance detection, and fuse multi-camera data to calculate and obtain the posture data.

[0024] Furthermore, in this embodiment, the execution layer includes an industrial robot, which is used to generate control instructions based on the posture data for transmission to the control end of the industrial robot, wherein, as shown in Figure 1, the computing layer sends control instructions to the robot control end through the EtherCAT bus, and the robot feeds back status information (such as joint angles, end posture) to the computing layer to form a closed-loop control.

[0025] Figure 2 A flow chart of an industrial robot optical navigation anti-occlusion tracking method of the present application is shown.

[0026] like Figure 2 As shown, the present application discloses an industrial robot optical navigation anti-occlusion tracking method, which is applied to any of the industrial robot optical navigation anti-occlusion tracking systems described above. The method comprises the following steps: S202, collecting multi-camera data and performing preprocessing to obtain a target image, wherein the multi-camera includes two active pan-tilt cameras and four fixed cameras; S204, performing pose calculation, motion prediction, and adaptive weighted fusion on the target image to obtain pose data; S206 , calculating target angles of joints of the robot based on the posture data to generate control instructions, thereby controlling the joint movements of the robot based on the control instructions.

[0027] It should be noted that in this embodiment, the workpiece is provided with a marking point. Therefore, during the movement of the workpiece, multi-camera data is collected and pre-processed to obtain a target image, and the image pre-processing is performed by Kalman filtering. The multi-camera includes two active pan-tilt cameras and four fixed cameras. Position constraints and posture constraints are defined to ensure that the visibility of the marking point of the workpiece in any posture is improved to 98%. Specifically, Figure 3 As shown, the position constraint calculation formula is as follows: ; in, is the coordinate of the workpiece marking point, is the equation of the j-th boundary plane of the Measurement Volume (MV), is the envelope radius of the marker point.

[0028] The posture constraint calculation formula is as follows: ; in, is the camera optical axis direction vector, is the normal vector of the marker point.

[0029] Furthermore, the target image is subjected to posture calculation, motion prediction and adaptive weighted fusion to obtain posture data, which will be described in detail in the subsequent instructions. Then, based on the posture data, the target angles of each joint of the robot are calculated to generate control instructions, thereby controlling the robot joint movements based on the control instructions. Real-time control communication is performed based on the EtherCAT bus, and the repeatability positioning accuracy is ±0.03mm.

[0030] According to an embodiment of the present invention, performing pose calculation and motion prediction on the target image specifically includes: Extracting the target image, wherein the target image includes a dedistorted image and 2D pixel coordinates; Calculate the workpiece's 6D pose based on 2D pixel coordinates and preset 3D model coordinates; Motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose, wherein the correction is specifically performed iteratively through the Newton method.

[0031] It should be noted that, in this embodiment, the target image is extracted, and the target image includes a dedistorted image and 2D pixel coordinates. The 6D pose of the workpiece is calculated based on the 2D pixel coordinates and the preset 3D model coordinates. Then, motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose. Specifically, KF-Newton composite prediction is used: ; in, is the state transition matrix, is the state vector, is the control input matrix, is the control input vector, is the process noise, is the state correction amount, is the Hessian matrix, which converges after 3 iterations, where the state vector Used to describe the system's posture, velocity and other states at time k, and control the input matrix Used to map the control input to the state space, the control input vector Corresponding to external control signals, such as robot joint instructions, process noise Used to describe the uncertainty in the system model, state correction It is used to iteratively optimize the increment of pose estimation and further construct the QP problem: ; in, is the weight matrix, Indicates minimization optimization of variables. Represents velocity increment, usually used for velocity or position correction in motion prediction, while is the velocity increment The transpose of is a positive definite weight matrix used to weight the velocity increments of different dimensions and control the importance of each component in the optimization process. c is the constraint matrix. is the transpose of the constraint matrix c, G is the constraint upper bound vector corresponding to the constraint matrix c, and h is the velocity increment The absolute value constraint upper bound is used to limit the speed increment The maximum variation range of the workpiece is obtained by adjusting the position of the workpiece and obtaining the predicted pose.

[0032] According to an embodiment of the present invention, adaptive weighted fusion to obtain pose data specifically includes: Calculating a visibility score of the workpiece for each camera based on the 6D pose and the predicted pose; The multi-camera viewing angles are adjusted based on the visibility scores, and the pose data are obtained by performing adaptive weighted fusion calculation based on the confidence levels of different cameras.

[0033] It should be noted that, in this embodiment, the visibility score of each camera to the workpiece is calculated, wherein the visibility score includes scoring based on distance, normal direction, and occlusion. The viewing angles of multiple cameras are adjusted based on the visibility score, and the camera with the optimal viewing angle is selected to control the rotation of the gimbal to ensure that at least one camera is always visible. If multiple cameras are visible, the pose data is obtained by performing adaptive weighted fusion calculation based on the confidence of different cameras, wherein the confidence weight premise needs to eliminate abnormal poses based on the Mahalanobis distance.

[0034] Figure 4 A block diagram of an industrial robot optical navigation anti-occlusion tracking system of the present invention is shown.

[0035] like Figure 4 As shown, the present invention discloses an industrial robot optical navigation anti-occlusion tracking system 40, including a memory 41 and a processor 42. The memory 41 includes an industrial robot optical navigation anti-occlusion tracking method program. When the industrial robot optical navigation anti-occlusion tracking method program is executed by the processor 42, the following steps are implemented: Collecting multi-camera data and preprocessing it to obtain the target image, where the multi-camera includes two active pan-tilt cameras and four fixed cameras; Performing pose calculation, motion prediction, and adaptive weighted fusion on the target image to obtain pose data; The target angle of each joint of the robot is calculated based on the posture data to generate a control instruction, thereby controlling the robot joint movement based on the control instruction.

[0036] It should be noted that in this embodiment, the workpiece is provided with a marking point. Therefore, during the movement of the workpiece, multi-camera data is collected and pre-processed to obtain a target image, and the image pre-processing is performed by Kalman filtering. The multi-camera includes two active pan-tilt cameras and four fixed cameras. Position constraints and posture constraints are defined to ensure that the visibility of the marking point of the workpiece in any posture is improved to 98%. Specifically, Figure 3 As shown, the position constraint calculation formula is as follows: ; in, is the coordinate of the workpiece marking point, is the equation of the j-th boundary plane of the Measurement Volume (MV), is the envelope radius of the marker point.

[0037] The posture constraint calculation formula is as follows: ; in, is the camera optical axis direction vector, is the normal vector of the marker point.

[0038] Furthermore, the target image is subjected to posture calculation, motion prediction and adaptive weighted fusion to obtain posture data, which will be described in detail in the subsequent instructions. Then, based on the posture data, the target angles of each joint of the robot are calculated to generate control instructions, thereby controlling the robot joint movements based on the control instructions. Real-time control communication is performed based on the EtherCAT bus, and the repeatability positioning accuracy is ±0.03mm.

[0039] According to an embodiment of the present invention, performing pose calculation and motion prediction on the target image specifically includes: Extracting the target image, wherein the target image includes a dedistorted image and 2D pixel coordinates; Calculate the workpiece's 6D pose based on 2D pixel coordinates and preset 3D model coordinates; Motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose, wherein the correction is specifically performed iteratively through the Newton method.

[0040] It should be noted that, in this embodiment, the target image is extracted, and the target image includes a dedistorted image and 2D pixel coordinates. The 6D pose of the workpiece is calculated based on the 2D pixel coordinates and the preset 3D model coordinates. Then, motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose. Specifically, KF-Newton composite prediction is used: ; in, is the state transition matrix, is the state vector, is the control input matrix, is the control input vector, is the process noise, State correction is the Hessian matrix, which converges after 3 iterations, where the state vector Used to describe the system's posture, velocity and other states at time k, and control the input matrix Used to map the control input to the state space, the control input vector Corresponding to external control signals, such as robot joint instructions, process noise Used to describe the uncertainty in the system model, state correction It is used to iteratively optimize the increment of pose estimation. Further, the QP problem is constructed: ; in, is the weight matrix, Indicates minimization optimization of variables. Represents velocity increment, usually used for velocity or position correction in motion prediction, while is the velocity increment The transpose of is a positive definite weight matrix used to weight the velocity increments of different dimensions and control the importance of each component in the optimization process. c is the constraint matrix. is the transpose of the constraint matrix c, G is the constraint upper bound vector corresponding to the constraint matrix c, and h is the velocity increment The absolute value constraint upper bound is used to limit the speed increment The maximum variation range of the workpiece is calculated, where the constraints include the robot arm joint limits. This is done to correct the workpiece motion position and obtain the predicted pose.

[0041] According to an embodiment of the present invention, adaptive weighted fusion to obtain pose data specifically includes: Calculating a visibility score of the workpiece for each camera based on the 6D pose and the predicted pose; The multi-camera viewing angles are adjusted based on the visibility scores, and the pose data are obtained by performing adaptive weighted fusion calculation based on the confidence levels of different cameras.

[0042] It should be noted that, in this embodiment, the visibility score of each camera to the workpiece is calculated, wherein the visibility score includes scoring based on distance, normal direction, and occlusion. The viewing angles of multiple cameras are adjusted based on the visibility score, and the camera with the optimal viewing angle is selected to control the rotation of the gimbal to ensure that at least one camera is always visible. If multiple cameras are visible, the pose data is obtained by performing adaptive weighted fusion calculation based on the confidence of different cameras, wherein the confidence weight premise needs to eliminate abnormal poses based on the Mahalanobis distance.

[0043] The fourth aspect of the present invention provides a computer-readable storage medium, which includes an industrial robot optical navigation anti-occlusion tracking method program. When the industrial robot optical navigation anti-occlusion tracking method program is executed by a processor, it implements the steps of an industrial robot optical navigation anti-occlusion tracking method as described in any one of the above items.

[0044] The present invention discloses an industrial robot optical navigation anti-occlusion tracking system, method, device and medium. The core innovations include: a normal direction constraint model: establishing dual constraints including position and posture, defining a composite optimization objective function of the normal direction angle α and the distance angle β, so that the visibility of the workpiece's marking points in any posture is improved to 98%; and a hierarchical motion prediction algorithm: combining Kalman filtering (KF) and Newton iteration method to reduce the 6D posture prediction delay from 120ms to 18ms, and increase the speed tolerance to 2.5m / s; and an adaptive field of view fusion strategy: through Mahalanobis distance test and confidence weighting, it still maintains a positioning accuracy of 0.8mm under 50% occlusion (the traditional method is 2.1mm).

[0045] Among them, an anti-occlusion test was carried out in the automobile door frame welding scenario. The working conditions were two moving workpieces and there was occlusion of the robot body. In the test results, the traditional method had an error of 2.2mm and a failure rate of 30% when occluded, while the maximum error of this application was 0.7mm and the failure rate was 3.4%. Experiments showed that the tracking failure rate of this system in the automobile welding scenario was reduced to 3.2%, which is 10 times higher than that of the traditional solution; power consumption was reduced by 40%, and the cost was only 1 / 4 of the commercial system.

[0046] Specifically, this application significantly improves the tracking accuracy and real-time performance of industrial robots under complex working conditions through dynamic viewpoint optimization, normal direction constraint modeling and hierarchical motion prediction algorithm, while reducing costs and power consumption, and has broad application prospects.

[0047] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0048] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0049] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0050] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0051] Alternatively, if the integrated units described above are implemented as software modules and sold or used as standalone products, they can also be stored on a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product, stored on a storage medium, includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as removable storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. An industrial robot optical navigation anti-occlusion tracking system, characterized in that: include: Perception layer, computing layer and execution layer, among which, The perception layer includes a hybrid vision network composed of two active pan-tilt cameras and four fixed cameras, and is used to obtain a global workpiece image of the workpiece, wherein the workpiece is provided with marking points; The computing layer includes a posture solving module, a motion prediction module and a fusion decision module, which are used to receive the image data of the perception layer and process it to obtain posture data; The execution layer includes an industrial robot, which is used to generate control instructions based on the posture data and transmit them to the control end of the industrial robot.

2. The industrial robot optical navigation anti-occlusion tracking system according to claim 1, characterized in that: The posture calculation module is used to identify the marked points in the workpiece image; the motion prediction module is used to correct the workpiece motion position; the fusion decision module is used to eliminate abnormal postures through Mahalanobis distance detection and fuse multi-camera data to calculate the posture data.

3. An industrial robot optical navigation anti-occlusion tracking method, characterized in that: The method for an industrial robot optical navigation anti-occlusion tracking system according to any one of claims 1 to 2 comprises the following steps: Collecting multi-camera data and preprocessing it to obtain the target image, where the multi-camera includes two active pan-tilt cameras and four fixed cameras; Performing pose calculation, motion prediction, and adaptive weighted fusion on the target image to obtain pose data; The target angle of each joint of the robot is calculated based on the posture data to generate a control instruction, thereby controlling the robot joint movement based on the control instruction.

4. The method for optical navigation and anti-occlusion tracking of an industrial robot according to claim 3, characterized in that: Performing pose calculation and motion prediction on the target image, specifically including: Extracting the target image, wherein the target image includes a dedistorted image and 2D pixel coordinates; Calculate the workpiece's 6D pose based on 2D pixel coordinates and preset 3D model coordinates; Motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose, wherein the correction is specifically performed iteratively through the Newton method.

5. The method for optical navigation and anti-occlusion tracking of an industrial robot according to claim 4, characterized in that: Adaptive weighted fusion obtains pose data, including: Calculating a visibility score of the workpiece for each camera based on the 6D pose and the predicted pose; The multi-camera viewing angles are adjusted based on the visibility scores, and the pose data are obtained by performing adaptive weighted fusion calculation based on the confidence levels of different cameras.

6. An industrial robot optical navigation anti-occlusion tracking device, characterized in that: The invention comprises a memory and a processor, wherein the memory comprises an industrial robot optical navigation anti-occlusion tracking method program, and the industrial robot optical navigation anti-occlusion tracking method program is executed by the processor to implement the following steps: Collecting multi-camera data and preprocessing it to obtain the target image, where the multi-camera includes two active pan-tilt cameras and four fixed cameras; Performing pose calculation, motion prediction, and adaptive weighted fusion on the target image to obtain pose data; The target angle of each joint of the robot is calculated based on the posture data to generate a control instruction, thereby controlling the robot joint movement based on the control instruction.

7. The optical navigation anti-occlusion tracking device for industrial robots according to claim 6, characterized in that: Performing pose calculation and motion prediction on the target image, specifically including: Extracting the target image, wherein the target image includes a dedistorted image and 2D pixel coordinates; Calculate the workpiece's 6D pose based on 2D pixel coordinates and preset 3D model coordinates; Motion prediction is performed based on the 6D pose of the workpiece to correct the motion position of the workpiece to obtain a predicted pose, wherein the correction is specifically performed iteratively through the Newton method.

8. The optical navigation anti-occlusion tracking device for industrial robots according to claim 7, characterized in that: Adaptive weighted fusion obtains pose data, including: Calculating a visibility score of the workpiece for each camera based on the 6D pose and the predicted pose; The multi-camera viewing angles are adjusted based on the visibility scores, and the pose data are obtained by performing adaptive weighted fusion calculation based on the confidence levels of different cameras.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes an industrial robot optical navigation anti-occlusion tracking method program. When the industrial robot optical navigation anti-occlusion tracking method program is executed by a processor, the steps of an industrial robot optical navigation anti-occlusion tracking method as described in any one of claims 3 to 5 are implemented.

Images