A method for calibrating installation pose of a depth camera based on two-dimensional code recognition

Abstract

A kind of depth camera installation pose calibration method based on two-dimensional code identification, the pose of the RGB camera of depth camera relative to two-dimensional code is calculated by two-dimensional code positioning algorithm, it is transformed into the pose between two-dimensional code relative to IR camera by transformation matrix between RGB camera and IR camera, the pose between robot and two-dimensional code is known, the pose between robot and IR camera is calculated, i.e. the pose required by depth camera. It can solve the installation error of depth camera, according to the two-dimensional code identified by RGB camera, the accurate pose relationship between mobile robot and depth camera is effectively calibrated, and the subsequent use of depth camera is more accurate.

Description

Technical Field

[0001] This invention relates to the field of mobile robot image processing technology, specifically to a method for depth camera mounting pose calibration based on QR code recognition. Background Technology

[0002] In the process of navigation and obstacle avoidance of mobile robots, depth cameras installed at a top-down angle are often used to supplement the blind spots of LiDAR. Due to the installation error of the depth camera, the installation parameters need to be calibrated. However, there is currently a lack of a fast and effective automatic calibration method for the installation pose of the depth camera. Summary of the Invention

[0003] To overcome the shortcomings of the above technologies, this invention provides a method for effectively calibrating the accurate pose relationship between a mobile robot and a depth camera.

[0004] The technical solution adopted by this invention to overcome its technical problems is:

[0005] A method for depth camera mounting pose calibration based on QR code recognition includes the following steps:

[0006] a) Equip the robot with a depth camera that looks diagonally downwards, and calibrate the intrinsic parameters of the RGB camera of the depth camera; b) Place a QR code on the ground at a specified distance in front of the robot;

[0007] c) Calculate the pose of the depth camera's RGB camera relative to the QR code on the ground, and obtain the rotation matrix R and translation vector T;

[0008] d) The transformation matrix between the RGB camera and the IR camera of the depth camera on the robot is R′, the translation vector between the RGB camera and the IR camera of the depth camera is T′, the rotation matrix between the QR code and the IR camera is RR′, and the translation vector between the QR code and the IR camera is T+T′.

[0009] e) Convert the translation vector T+T′ to coordinate values ​​under base_link, and convert the rotation matrix RR′ to rotation angle under base_link to obtain the mounting pose of the depth camera.

[0010] Furthermore, in step a), the RGB camera is calibrated using the Zhang Zhengyou calibration method to obtain the intrinsic parameter matrix of the depth camera's RGB camera. Where f x f is the focal length of the depth camera in pixels along the X-axis. y denoted as the focal length of the depth camera in pixels along the Y-axis, u0 is the horizontal pixel coordinate of the optical center in the image, and v0 is the vertical pixel coordinate of the optical center in the image.

[0011] Furthermore, in step c), the p4p algorithm is used to calculate the pose of the depth camera's RGB camera relative to the QR code.

[0012] The beneficial effects of this invention are as follows: The pose of the depth camera relative to the QR code is calculated using a QR code positioning algorithm. This pose is then transformed into the pose of the QR code relative to the IR camera using a transformation matrix between the RGB and IR cameras. Since the pose between the robot and the QR code is known, calculating the pose between the robot and the IR camera yields the required pose of the depth camera. This invention can resolve depth camera installation errors and, by recognizing the QR code using the RGB camera, effectively calibrate the accurate pose relationship between the mobile robot and the depth camera, leading to more precise use of the depth camera in subsequent applications. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the QR code of the present invention;

[0014] Figure 2 This is a schematic diagram of the pose calibration method of the present invention. Detailed Implementation

[0015] The following is in conjunction with the appendix Figure 1 Appendix Figure 2 The present invention will be further described below.

[0016] A method for depth camera mounting pose calibration based on QR code recognition includes the following steps:

[0017] a) The robot is equipped with a depth camera that looks obliquely downwards, and the intrinsic parameters of the RGB camera of the depth camera are calibrated.

[0018] b) Place a QR code on the ground at a specified distance in front of the robot.

[0019] c) Calculate the pose of the depth camera's RGB camera relative to the QR code on the ground, and obtain the rotation matrix R and translation vector T.

[0020] d) The transformation matrix between the RGB camera and the IR camera of the depth camera on the robot is R′, the translation vector between the RGB camera and the IR camera of the depth camera is T′, the rotation matrix between the QR code and the IR camera is RR′, and the translation vector between the QR code and the IR camera is T+T′.

[0021] e) Convert the translation vector T+T′ to coordinate values ​​under base_link, and convert the rotation matrix RR′ to rotation angle under base_link to obtain the mounting pose of the depth camera.

[0022] The robot is equipped with a depth camera that looks downwards at an angle. An RGB camera is calibrated, and a QR code of a specified size and format is placed on the ground at a designated distance in front of the robot. A QR code localization algorithm calculates the pose of the depth camera's RGB camera relative to the QR code. This pose is then transformed into the pose of the QR code relative to the IR camera using a transformation matrix between the RGB and IR cameras. Since the pose between the robot and the QR code is known, the pose between the robot and the IR camera is calculated, which yields the required pose for the depth camera. This method can resolve depth camera installation errors and, by recognizing the QR code using the RGB camera, effectively calibrates the accurate pose relationship between the mobile robot and the depth camera, resulting in more precise use of the depth camera in subsequent applications.

[0023] In one embodiment of the present invention, in step a), the RGB camera is calibrated using the Zhang Zhengyou calibration method to obtain the intrinsic parameter matrix of the RGB camera of the depth camera. Where f x f is the focal length of the depth camera in pixels along the X-axis. y Let u0 be the focal length of the depth camera in pixels along the Y-axis, and v0 be the horizontal pixel coordinate of the optical center in the image. In one embodiment of the invention, step c) uses the p4p algorithm to calculate the pose of the RGB camera of the depth camera relative to the QR code.

[0024] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for calibrating a mounting pose of a depth camera based on two-dimensional code recognition, characterized in that, Comprising the following steps: a) the robot sets a depth camera looking downward to the dip slope, calibrates the intrinsic parameters of the RGB camera of the depth camera; b) place a two-dimensional code on the ground in front of the robot at a specified distance; c) calculating the pose of the RGB camera of the depth camera relative to the two-dimensional code on the ground by means of the two-dimensional code, resulting in a rotation matrix and a translation vector ; d) the transformation matrix between the RGB camera and the IR camera of the depth camera on the robot is , the translation vector between the RGB camera and the IR camera of the depth camera is , the rotation matrix between the QR code and the IR camera is , the translation vector between the QR code and the IR camera is ; e) converting the translation vector into coordinate values under base_link, converting the rotation matrix into rotation angles under base_link, obtaining the installation pose of the depth camera, wherein the pose between the robot and the two-dimensional code is known.

2. The method of calibrating the installation pose of a depth camera based on two-dimensional code recognition according to claim 1, characterized in that: In step a), calibrate the RGB camera by Zhang Zhengyou calibration method to obtain the intrinsic matrix of the RGB camera of the depth camera wherein is the focal length of the depth camera in the X-axis direction in units of pixels, is the focal length of the depth camera in the Y-axis direction in units of pixels, is the horizontal pixel coordinate of the optical center on the image, is the vertical pixel coordinate of the optical center on the image.

3. The method of claim 1, wherein: In step c), the pose of the RGB camera of the depth camera relative to the two-dimensional code is calculated using the p4p algorithm.

Images