CN110842928B - Visual guiding and positioning method for compound robot - Google Patents

Abstract

The invention provides a visual guidance positioning method for a compound robot. By recording the relative position and attitude information of the compound robot in detail during calibration, including the distance, position and angle from the workbench, the relative position can be reproduced during operation and application. It can cover the visual and action deviation of the composite robot due to the navigation and positioning error of the trolley and the unequal ground, so as to improve the operation accuracy of the robot in the actual environment. The present invention ensures that the operation accuracy of the compound robot is not affected by various actual error factors through accurate relative positioning based on various sensor information.

Description

A Vision-Guided Localization Method for Composite Robots

technical field

The present disclosure belongs to the field of industrial robots, in particular to a robot guidance and positioning technology, and in particular relates to a compound robot vision guidance and positioning method.

Background technique

The composite robot is a mobile manipulator composed of an AGV (automatic mobile vehicle) and a mechanical arm. It belongs to a new type of robot and can be quickly deployed in the 3C industry, automated factories, warehouse sorting, and automated goods supermarkets to realize automatic material handling, item loading and unloading, Material sorting, etc.

At present, the vision-guided positioning of the composite robot is the same as that of the fixed-base robot, and the eye-on-hand method is mainly adopted, that is, the end of the mechanical arm drives the camera to move. However, the working environment of the composite robot is more complicated than that of the fixed-base robot, which leads to this kind of robot. The difficulty of grasping and placing objects for visual guidance also increases.

The working area facing the base-fixed robot is fixed. Usually, the corresponding relationship between the hand and eye positions of each collection point is established through prior calibration, and the calibration parameters are calculated. Subsequently, only the pixel coordinates and calibration parameters of the object in the field of view need to be calculated. The displacement of the manipulator can be obtained, and the movement of the manipulator can be controlled according to the displacement, so that the workpiece placement can be accurately aligned, thereby achieving accurate pick and place. It can be seen that before the robotic arm picks up and places the workpiece, the alignment with the workpiece position can be achieved only by calculation and adjustment in the two-dimensional plane.

The composite robot with moving base needs to automatically navigate the AGV car to the parking spot, and then use the robotic arm to pick and place objects at the target position on the workbench. During this process, the parking spot of the AGV car will exist. Errors, as well as the unevenness of the ground at the parking spot of the AGV car, will cause the position and posture of the end of the robotic arm relative to the worktable to be different after the car reaches the parking spot. Since the robotic arm is a high-precision device, if you simply follow The processing method of the base-fixed robot, that is, only calculating the offset of the robot arm based on the calibration parameters and the pixel coordinates of the target position in the camera's field of view, will cause the end position of the robot arm to deviate from the actual target position, and the movement accuracy of the robot arm will be greatly reduced. .

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a visual guidance and positioning device and method for a compound robot, so as to solve the problem that the precision of the visual guidance of the compound robot decreases in the actual working environment.

One aspect of the present invention provides a compound robot vision guidance and positioning device, comprising:

The measurement module is fixed on the end of the manipulator, and is used to measure the distance of the end of the manipulator relative to the worktable, the inclination of the end of the manipulator, and obtain the image of the identification area on the worktable as a guide reference;

The vision guidance processing module is used to control the motion of the manipulator through the robot motion controller according to the information collected by the measurement module, so that the robot can make the distance, position and angle of the end of the manipulator relative to the worktable during calibration and work application. The value reaches the same state; control the automatic collection of calibration data and the calculation of calibration parameters; control the end of the manipulator to shift from the calibration position along the TCP to the workpiece placement position to complete the workpiece pick and place.

Further, the measurement module includes a ranging sensor, an inclination sensor and an industrial camera.

Further, the visual guidance processing module includes:

Measurement module communication unit: communicates with the measurement module, acquires and records the distance and inclination information collected by the measurement module in real time;

Robot communication unit: send commands to the robot motion controller, instruct it to control the robot motion, and obtain the coordinate information of the robot in real time;

Image recognition and positioning unit: control the measurement module to collect images, identify and position the received images, acquire and record the pixel coordinates of the reference points and the Rz angle values of the images;

Calibration parameter calculation unit: Calculate and save the calibration parameters according to the pixel coordinates of the reference point obtained at each collection point and the offset of the robotic arm;

Human-computer interaction unit: provides the user with the input and output interface of instructions, data, images and other information.

Further, it also includes a two-dimensional code label or a character label fixed to the identification area on the workbench.

The present invention also provides a vision-guided positioning method for a compound robot suitable for the above-mentioned device, including a step of vision-guided calibration and a step of vision-guided job application, wherein the step of vision-guided calibration specifically includes:

Relative pose acquisition: When the car reaches the appropriate parking point, adjust the end of the manipulator to the appropriate calibration origin, acquire and record the pose of the manipulator, the inclination of the end of the manipulator, and the distance, position and Rz relative to the work surface angle value;

Calibration parameter calculation: calibration data acquisition, calculation of calibration parameters according to the collected data;

Measure the TCP offset of the robot arm from the calibration origin to the workpiece placement position;

The steps of the vision-guided job application specifically include:

Recurrence of relative pose: When the car stops at the position and angle at the time of calibration, control the robotic arm to move to the pose recorded during calibration, and then control the end of the robotic arm to adjust to the relative pose recorded during calibration;

According to the TCP offset from the calibration origin to the workpiece placement position, the movement of the robotic arm is controlled to complete the pick-and-place task.

Further, the step of collecting the relative pose specifically includes:

Adjust the posture of the end of the manipulator so that the working plane is parallel to the work surface, and collect and record the values of the inclination angles Rx and Ry of the end of the manipulator;

Adjust the distance between the end of the robotic arm and the workbench to make the imaging field of view of the measurement module clear, and record the current point as the calibration origin;

Collect and record the distance from the end of the robotic arm to the work surface at this time;

Collect the image of the identification area on the workbench, identify and locate the image, extract and record the pixel coordinates of the reference point and the Rz angle value of the identification area;

Collect and record the pose of the robotic arm at this time.

Further, the method for identifying and positioning the image is as follows: using a shape matching method for identification and positioning, and if the identification area is provided with a two-dimensional code label, then using a two-dimensional code identification and positioning method for identification and positioning.

Further, the method for the calibration data collection includes:

Set the TCP offset of each collection point relative to the calibration origin;

Control the robotic arm to move to each collection point according to the preset offset along the TCP coordinate system;

The image of the identification area on the workbench is collected at each collection point, the image is identified and positioned, and the pixel coordinates of the reference point therein are extracted and recorded.

Further, the method for calculating the calibration parameters is:

Assuming that the end of the robot arm is at the calibration origin, the pixel coordinates of the reference point in the identification area are (u ₀ , v ₀ ); the end of the robot arm moves along the TCP coordinate system by Δx, and the pixel coordinates of the reference point are (u ₁ , v ₁ ); the end of the robotic arm returns to the calibration origin and then moves Δy along the TCP coordinate system, and the pixel coordinates of the reference point are (u ₂ , v ₂ );

Then in the job application, if the pixel coordinates of the reference point are (u, v), then the TCP offset (Δx', Δy') of the robot arm relative to the calibration origin should be:

in,

Δu ₁ =u ₁ -u ₀ , Δv ₁ =v ₁ -v ₀ , Δu ₂ =u ₂ -u ₀ , Δv ₂ =v ₂ -v ₀ , Δu=uu ₀ , Δv=vv ₀ .

Further, the step of reproducing the relative pose specifically includes:

Control the movement of the robotic arm to the posture recorded when it is calibrated;

Measure the values of the inclination angle Rx and Ry of the end of the manipulator, and adjust the posture of the end of the manipulator so that the inclination angle is the same as the value of the inclination angle recorded during calibration;

Measure the distance between the end of the robotic arm and the work surface, and adjust the posture of the robotic arm so that the distance is equal to the distance value recorded during calibration;

Obtain the image of the marking area on the workbench, extract the Rz angle value of the marking area, compare it with the image Rz value recorded during calibration, and adjust the attitude Rz of the end of the robotic arm to make the two consistent;

Obtain the image of the identification area again, obtain the pixel coordinates of the reference point in it, calculate the TCP offset of the robotic arm according to the recorded calibration parameters, and control the movement of the robotic arm according to the offset, so as to reach the relative value recorded during calibration. position on the work surface.

It can be seen that the visual guidance and positioning device and method of the compound robot provided by the present invention can reproduce the relative position and attitude information of the compound robot during calibration, including the distance, position and angle from the workbench, etc. A relative pose, covering the visual and action deviations of the compound robot due to the navigation and positioning error of the trolley and the unequal ground, thereby improving the operation accuracy of the robot in the actual environment. Compared with the prior art, its beneficial effects mainly include: (1) ensuring the accuracy of picking and placing objects of the compound robot in various unfavorable working environments; (2) being able to communicate with the robot motion controller to directly control the robot motion, There is no need to use the robotic arm teach pendant, which is convenient and fast, the data is accurate, and there is no human input error; (3) automatically identify and obtain the pixel coordinates and image angles of the vision guidance benchmark, reduce the operational complexity of the calibration process, and improve the calibration efficiency. .

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not intended to limit the scope of what is claimed in this application.

Description of drawings

The accompanying drawings, which are part of the present specification, illustrate embodiments of the invention, and together with the description, serve to explain the principles of the invention.

1 is a schematic diagram of the composition of an exemplary embodiment of a compound robot vision-guided positioning device;

2 is an overall flow chart of an exemplary embodiment of a compound robot vision-guided positioning method;

3 is a flow chart of calibration steps of an exemplary embodiment of a compound robot vision-guided positioning method;

FIG. 4 is a flow chart of operation application steps of an exemplary embodiment of a compound robot vision-guided positioning method.

Detailed ways

Preferred embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While preferred embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

FIG. 1 is a schematic composition diagram of an exemplary embodiment of the vision-guided positioning device for a compound robot according to the present disclosure.

The vision-guided positioning device for a compound robot provided by the present disclosure includes:

The measurement module is used to measure the distance of the end of the manipulator relative to the work table, the inclination of the end of the manipulator, and obtain the image of the identification area on the workbench as a guide reference;

The vision guidance processing module is used to control the motion of the manipulator through the robot motion controller according to the information collected by the measurement module, so that the robot can make the distance, position and angle of the end of the manipulator relative to the worktable during calibration and work application. The value reaches the same state; control the automatic collection of calibration data and the calculation of calibration parameters; control the end of the manipulator to shift from the calibration position along the TCP to the workpiece placement position to complete the workpiece pick and place.

Among them, the identification area on the work surface as a guide reference should have features that are obviously different from the surrounding environment and are easy to be identified and distinguished. A characteristic point is selected as the reference point, and the positioning of the end of the robot arm relative to the worktable is realized by the pixel coordinate positioning of this point in the image.

The position of the identification area can be selected flexibly: if the workpiece to be picked and placed is placed in the workpiece positioning tool on the worktable, and the position and posture of the workpiece relative to the positioning tool remain unchanged, then it can be directly selected on the workpiece positioning tool. The relative positioning relationship between the workpiece placement, workpiece positioning tooling, and identification area remains unchanged; if the position of the target workpiece to be grasped on the worktable is active, that is, there is no workpiece positioning tooling limit, then it can also be directly Select on the target workpiece to be grabbed. At the same time, it is necessary to select a position that is convenient for the camera device at the end of the robotic arm to identify.

The internal features of the identification area can be set by printing or sticking labels, and the preferred solution is to stick a two-dimensional code label or a character label.

The measurement module provides more information for measuring the position and attitude of the robot relative to the worktable, and then adjusts the position and attitude based on this, which can effectively correct the visual deviation of the compound robot due to the navigation and positioning error of the car and the unequal ground, and provide vision Guide job accuracy.

The measurement module can adopt all existing detection devices suitable for measuring the distance of the end of the manipulator relative to the worktable, the inclination of the end of the manipulator, and acquiring the image of the identification area on the worktable as a guide reference.

The preferred solution is a ranging sensor, an inclination sensor and an industrial camera, as shown in Figure 1.

The vision guidance processing module usually includes the robot's vision control computer and the vision guidance processing software running in it. The vision control computer, also known as the vision controller, is an industrial control computer. It has a communication interface with the robot motion controller and measurement module: it is generally connected to the robot motion controller through a network cable, and communicates through the TCP/IP protocol; The camera is connected; it is connected to the angle measuring or ranging sensor through the serial port. The visual guidance processing software completes various functions of the visual guidance processing module.

As a preferred solution, the visual guidance processing module includes:

Measurement module communication unit: communicates with the measurement module, acquires and records the distance and inclination information collected by the measurement module in real time;

Robot communication unit: send commands to the robot motion controller, instruct it to control the robot motion, and obtain the coordinate information of the robot in real time;

Image recognition and positioning unit: control the measurement module to collect images, identify and position the received images, acquire and record the pixel coordinates of the reference points and the Rz angle values of the images;

Calibration parameter calculation unit: Calculate and save the calibration parameters according to the pixel coordinates of the reference point obtained at each collection point and the offset of the robotic arm;

Human-computer interaction unit: provides the user with the input and output interface of instructions, data, images and other information.

FIG. 2 is a general flow chart of an exemplary embodiment of a vision-guided positioning method applicable to the above-mentioned compound robot vision-guided device. As shown in FIG. 2 , the vision guidance method according to the exemplary embodiment mainly includes the step of vision guidance calibration and the step of vision guidance job application, wherein the vision guidance calibration step includes:

Relative pose acquisition: After the car reaches the appropriate parking point, adjust the end of the manipulator to the appropriate calibration origin, acquire and record the pose of the manipulator, the inclination of the end of the manipulator, as well as the distance, position and Rz angle values relative to the work surface . Wherein, the position of the end of the robot arm relative to the worktable and the Rz angle value can be located by the pixel coordinates of the reference point in the identification area on the worktable as the guide reference and the Rz angle value of the identification area.

Calibration parameter calculation: including calibration data acquisition, and calculation of calibration parameters according to the collected data.

Measure the TCP offset of the robot arm from the calibration origin to the workpiece placement position.

Vision-guided job application steps include:

Relative pose reproduction: After the trolley moves to the parking point recorded during calibration, control the robotic arm to move to the pose recorded during calibration, and control the end of the robotic arm to adjust to the pose relative to the workbench during calibration according to the information collected by the measurement module .

According to the TCP offset from the calibration origin to the workpiece placement position obtained during calibration, the robotic arm is controlled to press the TCP offset to complete the pick-and-place task.

In this method, in the process of relative pose acquisition, through the acquisition of parameters obtained by the measurement module, the posture of the manipulator during calibration and the pose of the end of the manipulator relative to the worktable, including distance, position and angle, are recorded in detail, among which The position is characterized by the pixel coordinates of the reference point in the image of the identification area, and the angle is characterized by the angle values of the RX and Ry directions recorded by the tilt sensor and the Rz angle value of the image of the identification area. The pixel coordinates of the reference point in the identification area image and the Rz angle value of the identification area image are obtained by identifying and positioning the identification area image obtained by the measurement module.

After that, in the operation application, through the repetition of the above-mentioned relative pose, the pose of the composite robot relative to the worktable is the same as the calibration time, thereby shielding the robot's vision and actions caused by the navigation and positioning error of the car and the unequal ground. The deviation ensures the working accuracy of the robot in the actual working environment.

As a preferred solution, the steps of relative pose acquisition specifically include:

Adjust the posture of the end of the manipulator so that the working plane is parallel to the work surface, and collect and record the values of the inclination angles Rx and Ry of the end of the manipulator;

Adjust the distance between the end of the robotic arm and the workbench to make the imaging field of view of the measurement module clear, and record the current point as the calibration origin;

Collect and record the distance from the end of the robotic arm to the work surface at this time;

Collect the image of the identification area on the workbench, identify and locate the image, extract and record the pixel coordinates of the reference point and the Rz angle value of the identification area;

Collect and record the pose of the robotic arm at this time.

As a preferred solution, the steps of reproducing the relative pose specifically include:

Control the movement of the robotic arm to the posture recorded when it is calibrated;

Measure the values of the inclination angle Rx and Ry of the end of the manipulator, and adjust the posture of the end of the manipulator so that the inclination angle is the same as the value of the inclination angle recorded during calibration;

Measure the distance between the end of the robotic arm and the work surface, and adjust the posture of the robotic arm so that the distance is equal to the distance value recorded during calibration;

Obtain the image of the marking area on the workbench, extract the Rz angle value of the marking area, compare it with the image Rz value recorded during calibration, and adjust the attitude Rz of the end of the robotic arm to make the two consistent;

Obtain the image of the identification area again, obtain the pixel coordinates of the reference point in it, calculate the TCP offset of the robotic arm according to the recorded calibration parameters, and control the movement of the robotic arm according to the offset, so as to reach the relative value recorded during calibration. position on the work surface.

Among them, the method of image recognition and positioning is a mature technology, and the method of shape matching is usually used, but if a two-dimensional code label is set in the identification area, the method of two-dimensional code recognition and positioning can be used.

All applicable methods in the prior art can be used for the calculation of the calibration parameters of the compound robot. The preferred solution is calibration data acquisition and calibration parameter calculation based on incremental compensation: according to the preset TCP offsets in the Rx and Ry directions, control the manipulator to reach each preset point around the calibration origin, and collect the workbench at this point. The pixel coordinates of the upper reference point are calculated, and the calibration parameters are calculated according to the collected data, and the corresponding relationship between the pixel coordinates of the reference point and the TCP offset of the robot arm relative to the calibration origin is established.

In addition, in the calibration step, the measurement of the TCP offset of the robot arm from the calibration origin to the workpiece placement position can be accomplished by means of traction teaching or the like.

The vision-guided positioning method of the compound robot of the present invention shields the influence of various error factors of the robot car itself through detailed relative pose measurement during calibration and repetition in the subsequent operation and application links, and ensures the workpiece picking and placing accuracy. Convenient and precise.

Application example:

Figure 1 shows an exemplary compound robot vision-guided positioning device. In the figure, the compound robot is composed of a mobile cart (1) and a mechanical arm (2), and the work surface (10) is in a horizontal state. When the trolley is calibrated or working, a suitable parking point (11) is selected, and the workpiece positions of the workbench are all within the working range (9) of the end of the mechanical arm.

The robot vision-guided positioning device includes:

The inclination sensor (3) installed horizontally at the end of the mechanical arm of the compound robot, the distance measuring sensor (4) installed at the end of the mechanical arm, and the industrial camera (5) are respectively used to measure the inclination of the end of the mechanical arm and the end of the mechanical arm to the workbench distance, and obtain the image of the identification area (7) as a visual guide reference;

The vision guidance processing module has communication interfaces with the robot motion controller, the inclination sensor (3), the ranging sensor (4), and the industrial camera (5), so as to obtain the distance, inclination, image and other data collected by the measurement module, Identify and locate the image, and obtain the pixel coordinates of the reference point and the Rz angle value of the identification area; send instructions to the robot motion controller to control the robot's motion, and at the same time obtain the coordinate information of the robot in real time.

The vision guidance processing module is mainly used to: control the offset of the manipulator according to the preset offset, collect the pixel coordinates of the reference point, and calculate the calibration parameters; acquire and record the distance, position and The angle value, control the end of the manipulator to reach the same relative pose when the job is applied; control the end of the manipulator to shift from the calibration position along the TCP to the workpiece placement position to complete the workpiece pick and place.

The vision guidance processing module includes the robot vision controller and the vision guidance processing software running in it. The vision controller is an industrial computer, which is respectively connected with the composite robot motion controller and the camera through the network cable, and is connected with the inclination sensor and the distance sensor through the serial port; Vision-guided processing software completes various specific functions.

In addition, the visual guidance positioning device also includes a two-dimensional code label or a character label arranged in the identification area (7), and uses the center point of the label as a reference point to perform pixel coordinate positioning.

The vision-guided positioning method for a compound robot suitable for the above-mentioned device includes a vision-guided positioning calibration step and an operation step, wherein:

S1, the calibration step, as shown in Figure 3, includes:

S11, after the composite robot AGV trolley arrives at the appropriate working parking point (the motion navigation of the trolley is responsible for the control device of the trolley), adjust the posture of the end of the composite robot manipulator so that the end of the manipulator is parallel to the worktable, and the visual guidance processing module records the moment. The angle values of the tilt sensors Rx and Ry.

S12, adjust the distance between the mechanical arm of the compound robot and the worktable, adjust the focal length and aperture of the industrial camera, so that the imaging field of view of the camera is clear, and the vision guidance processing module records the distance value of the ranging sensor at this moment. Take this point as the calibration origin.

S13, the vision guidance processing module communicates with the robot, records the pose of the robot arm at the moment, turns on the camera to capture images, locates the pixel coordinates and angle values of the label and records them.

S14, set the offset of the composite robot manipulator along the X direction and the Y direction of the TCP coordinate system (tool coordinate system), respectively control the manipulator to move to the offset point, turn on the camera to capture images, and perform pixel coordinate positioning on the label; The arm returns to the calibration origin, and then controls the robot to offset to another point along the TCP, turns on the camera to capture images, and performs pixel positioning on the labels in the image, and the robotic arm returns to the calibration origin.

S15, the vision guidance processing module calculates the calibration parameters according to the obtained pixel coordinates and the offset of the manipulator, and saves the calibration results locally.

S16 , move the robotic arm from the current calibration origin position along the TCP to the target workpiece position to be picked and place by means of traction teaching, and record the TCP offset (ΔTx', ΔTy', ΔTz') of the robotic arm.

S2, the operation steps, as shown in Figure 4, include:

S21, after the composite robot AGV car navigates from the initial point to the parking point during the calibration, moves the robotic arm to the calibration origin attitude;

S22, adjust the posture of the end of the manipulator according to the values of the inclination sensors Rx and Ry recorded during the calibration to ensure that the values of the inclination sensors are the same as those recorded during the calibration;

S23, adjust the distance from the end of the mechanical arm to the workbench according to the distance of the distance measuring sensor recorded during calibration;

S24, turn on the industrial camera to take a picture, identify the angle of the label, compare it with the Rz recorded during calibration, and send the angle deviation to the mechanical arm of the compound robot;

S25, the composite robot manipulator adjusts the end posture Rz according to the angle deviation to make it consistent with the posture during calibration;

S26, turn on the camera again to take a picture, identify the pixel coordinates of the positioning label, calculate the TCP offset of the robotic arm according to the calibration parameters, and command and control the robotic arm to perform the TCP offset. At this moment, the position and angle of the tool coordinate system of the composite robot manipulator arm relative to the label have been consistent with the calibration time.

S27, command and control the robotic arm to move according to the TCP according to the TCP offset of the robotic arm from the calibration origin to the position of the workpiece to be grasped recorded during the calibration. At this moment, the robotic arm can accurately run to the position of the workpiece to be picked and placed to complete the task of picking and placing.

It can be seen that the visual guidance and positioning device and method of the compound robot in this embodiment record the distance, position and angle of the robot arm relative to the worktable in detail by acquiring the angle measurement, distance measurement sensor data and camera images in real time, and in the subsequent actual During the work process, the robotic arm is precisely adjusted to achieve this relative pose, thereby shielding the impact of the positioning error, angle error, and ground inequality on the grasping and placing operation of the robotic arm due to the parking position error of the trolley, enabling the compound robot to visually guide the robot. The positioning accuracy can be significantly improved compared to the prior art.

The above are only illustrative specific embodiments of the present application. Without departing from the concept and principles of the present application, any equivalent changes and modifications made by those skilled in the art shall fall within the protection scope of the present application. .

Claims (6)

1. A vision-guided positioning method for a composite robot, comprising a step of vision-guided calibration and a step of vision-guided job application, wherein a measurement module is fixed at the end of the robotic arm of the composite robot for measuring the distance, The inclination of the end of the robot arm, and the image of the identification area on the workbench as a guide reference; The steps of the vision-guided calibration specifically include: Relative pose acquisition: After the car reaches the appropriate parking point, adjust the end of the manipulator to the appropriate calibration origin, acquire and record the pose of the manipulator, the inclination of the end of the manipulator, the distance relative to the work surface, the position and the marking area. Rz angle value; Calibration parameter calculation: calibration data acquisition, calculation of calibration parameters according to the collected data; Measure the TCP offset of the robotic arm from the calibration origin to the workpiece placement position; The steps of the vision-guided job application specifically include: Recurrence of relative pose: When the car stops at the position and angle at the time of calibration, control the robotic arm to move to the pose recorded during calibration, and then control the end of the robotic arm to adjust to the relative pose recorded during calibration; According to the TCP offset from the calibration origin to the workpiece placement position, the movement of the robotic arm is controlled to complete the pick-and-place task. 2. The compound robot vision-guided positioning method according to claim 1, wherein the step of collecting the relative pose specifically comprises: Adjust the posture of the end of the manipulator so that the working plane is parallel to the work surface, and collect and record the values of the inclination angles Rx and Ry of the end of the manipulator; Adjust the distance between the end of the robotic arm and the workbench to make the imaging field of view of the measurement module clear, and record the current point as the calibration origin; Collect and record the distance from the end of the robotic arm to the work surface at this time; Collect the image of the identification area on the workbench, identify and locate the image, extract and record the pixel coordinates of the reference point and the Rz angle value of the identification area; Collect and record the pose of the robotic arm at this time. 3. The compound robot vision-guided positioning method according to claim 2, wherein the method for identifying and positioning the image is: using a shape matching method to identify and position, and if the identification area is provided with a two-dimensional code label , the two-dimensional code identification and positioning method is used for identification and positioning. 4. The compound robot vision-guided positioning method according to claim 1, wherein the method for collecting the calibration data comprises: Set the TCP offset of each collection point relative to the calibration origin; Control the manipulator to move to each collection point according to the preset offset along the TCP coordinate system; The image of the identification area on the workbench is collected at each collection point, the image is identified and positioned, and the pixel coordinates of the reference point therein are extracted and recorded. 5. The compound robot vision-guided positioning method according to claim 4, wherein the method for calculating the calibration parameters is: Assuming that the end of the robot arm is at the calibration origin, the pixel coordinates of the reference point in the identification area are (u ₀ , v ₀ ); the end of the robot arm moves along the TCP coordinate system by Δx, and the pixel coordinates of the reference point are (u ₁ , v ₁ ); the end of the robotic arm returns to the calibration origin and then moves Δy along the TCP coordinate system, and the pixel coordinates of the reference point are (u ₂ , v ₂ ); Then in the job application, if the pixel coordinates of the reference point are (u, v), then the TCP offset (Δx', Δy') of the robot arm relative to the calibration origin should be:

in, Δu ₁ =u ₁ -u ₀ , Δv ₁ =v ₁ -v ₀ , Δu ₂ =u ₂ -u ₀ , Δv ₂ =v ₂ -v ₀ , Δu=uu ₀ , Δv=vv ₀ . 6. The compound robot vision-guided positioning method according to claim 1, wherein the step of reproducing the relative pose specifically comprises: Control the movement of the robotic arm to the posture recorded when it is calibrated; Measure the values of the inclination angle Rx and Ry of the end of the manipulator, and adjust the posture of the end of the manipulator so that the inclination angle is the same as the value of the inclination angle recorded during calibration; Measure the distance between the end of the robotic arm and the work surface, and adjust the posture of the robotic arm so that the distance is equal to the distance value recorded during calibration; Obtain the image of the marking area on the workbench, extract the Rz angle value of the marking area, compare it with the Rz value of the image recorded during calibration, and adjust the attitude Rz of the end of the robotic arm to make the two consistent; Obtain the image of the identification area again, obtain the pixel coordinates of the reference point in it, calculate the TCP offset of the robotic arm according to the recorded calibration parameters, and control the movement of the robotic arm according to the offset, so as to reach the relative value recorded during calibration. position on the work surface.

Images