Calibration method and calibration device

Abstract

The invention discloses a calibration method and a calibration device.The calibration method comprises the following steps that the calibration device used for internal reference calibration is provided, the calibration device comprises a center calibration plate and a plurality of side calibration plates arranged around the center calibration plate, and in the multiple side calibration plates, the center calibration plate and the side calibration plates are arranged in parallel; the unfolding angle and the size of each side calibration plate can be adjusted; adjusting the unfolding angles and / or the sizes of the side calibration plates based on the field angle of the camera module; and enabling the camera module to shoot the adjusted calibration image of the calibration device for calibration. The calibration device adopted by the invention can effectively solve the problem of insufficient corner coverage rate caused by limitation of the field angle of the camera module in a traditional calibration method, can adapt to the camera modules with different field angles, and aims at the camera modules with different field angles, such as ADAS and AVM. The flexible adjustment of the side calibration plate can improve the coverage rate of angular points in a calibration image, thereby remarkably improving the precision of internal reference calibration.

Description

Technical Field

[0001] The present invention relates to the field of camera technology, and in particular to a calibration method and a calibration device. Background Art

[0002] With the rapid development of intelligent automobile technology, the application of vehicle camera modules such as ADAS (Advanced Driver Assistance System) camera modules and AVM (Around View System) camera modules is becoming more and more widespread. These camera modules play a vital role in the vehicle driving process, such as providing lane departure warning, pedestrian detection, vehicle recognition, 360-degree panoramic imaging and other functions, which have a significant effect on improving driving safety and driving experience.

[0003] Intrinsic calibration is an indispensable part of the camera module calibration process. Its accuracy directly affects the quality of subsequent camera imaging, especially the effect of distortion removal. As the key feature points in the calibration process, the coverage rate of corner points directly affects the accuracy of intrinsic calibration. Traditional intrinsic calibration methods mostly use plane calibration plates, but they are limited by the size of the camera module's own FOV (field of view). Since the camera module is affected by the size of the lens's own FOV, the corner points often cannot completely cover the edge of the calibration plate. This problem is particularly prominent when dealing with camera modules with larger field of view angles such as AVM. When the AVM camera module shoots a plane calibration plate, the coverage rate of the intrinsic calibration corner points is generally around 60%; when the ADAS camera module shoots a plane calibration plate, the coverage rate of the intrinsic calibration angles is generally around 85%; and as the FOV of the camera module increases, the coverage rate of the captured corner points will decrease. In addition, traditional calibration methods often require taking multiple calibration images (such as more than 10 images) to improve the calibration accuracy, which not only increases the complexity and time cost of the calibration process, but may also introduce additional errors due to inconsistencies between images. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is to provide a calibration method and a calibration device that can improve the accuracy and efficiency of internal parameter calibration.

[0005] In order to solve the above technical problems, a technical solution adopted by the present invention is: to provide a calibration method, comprising the following steps:

[0006] A calibration device for internal reference calibration is provided, wherein the calibration device comprises a central calibration plate and a plurality of side calibration plates arranged around the central calibration plate, wherein each of the plurality of side calibration plates can adjust an expansion angle and its own size;

[0007] Adjusting the deployment angles and / or sizes of the plurality of side calibration plates based on the field of view of the camera module;

[0008] The camera module is enabled to capture the adjusted calibration image of the calibration device for calibration.

[0009] Furthermore, the central calibration plate is configured as a rectangular central calibration plate, and the multiple side calibration plates are configured as four side calibration plates respectively arranged on the four sides of the central calibration plate; among the four side calibration plates, the angle between each side calibration plate and the central calibration plate can be adjusted, and the size of each side calibration plate itself can be adjusted.

[0010] Furthermore, each of the multiple side calibration plates includes a mother light source plate and at least one child light source plate, wherein the at least one child light source plate can be received in the mother light source plate, and the at least one child light source plate can extend out of the mother light source plate from at least one side surface of the mother light source plate; and the central calibration plate is defined as a single light source plate.

[0011] Further, the number of sub-light source boards included in each side marking board is limited to two, and both sub-light source boards can be received in the mother light source board, and the two sub-light source boards can extend out of the mother light source board from opposite sides of the mother light source board;

[0012] All light source boards have the same size, and all light source boards are displayed with a chart for calibration and the specifications of the chart on all light source boards are the same. The light source boards include a mother light source board, a sub-light source board, and a single light source board.

[0013] Furthermore, the step of adjusting the calibration device according to the field of view of the camera module specifically includes:

[0014] If the field of view angle of the camera module is less than or equal to the first threshold, the side calibration plates of the calibration device are adjusted to unfold into a flat calibration plate; if the field of view angle of the camera module is greater than the first threshold and less than or equal to the second threshold, the calibration device is adjusted to a calibration light box with an opening at the bottom; if the field of view angle of the camera module is greater than the second threshold, the calibration device is adjusted to a calibration light box with an open bottom and the degree of opening is adjusted according to the specific field of view to adapt to the field of view angle.

[0015] Furthermore, the step of adjusting the side calibration plates of the calibration device to unfold them into a plane calibration plate specifically includes: adjusting the angle of the mother light source board of each side calibration board so that it is in the same horizontal plane as the single light source board, and extending a sub-light source board from each mother light source board, and all the light source boards are spliced ​​into a plane calibration plate.

[0016] The step of adjusting the calibration device to a calibration light box with a bottom opening specifically includes: controlling all sub-light source panels to be retracted into the main light source panel, adjusting the angle of each main light source panel to make it perpendicular to the single light source panel, and forming a calibration light box with a bottom opening.

[0017] Furthermore, the step of adjusting the calibration device to a calibration light box with an open bottom and adjusting the degree of opening thereof according to a specific field of view angle to adapt to the field of view angle specifically includes the following sub-steps:

[0018] Randomly adjust the included angle between each mother light source panel and the single light source panel to be the same and greater than 90 degrees, and extend the sub-light source panels so that the sub-light source panels extended from adjacent mother light source panels abut against each other;

[0019] Placing the camera module at the center of the calibration device, photographing the calibration device at this time, and obtaining a calibration image;

[0020] Processing the calibration image using an adaptive corner point algorithm to obtain a corner point coverage rate, and determining whether the corner point coverage rate is greater than or equal to a set third threshold;

[0021] If the corner point coverage is greater than or equal to a set third threshold, the calibration device is no longer adjusted;

[0022] If the corner point coverage is less than the set third threshold, repeat the above steps and continue to adjust the calibration device until the corner point coverage is greater than or equal to the set third threshold.

[0023] Further, the step of enabling the camera module to capture the adjusted calibration image of the calibration device for calibration specifically includes the following sub-steps:

[0024] Using an adaptive corner point algorithm to process the calibration image;

[0025] Intrinsic parameter calibration is performed based on the corner point data obtained after processing using the adaptive corner point algorithm.

[0026] Furthermore, the step of processing the calibration image using an adaptive corner point algorithm specifically includes the following sub-steps:

[0027] Remove background interference, extract the RGB color channel information of the calibration image, and ignore the Alpha channel;

[0028] Corner detection: detect the corner points on the calibration image after removing interference;

[0029] Calculate the area by connecting the outermost corner points detected to form a polygonal area, and calculate the area of ​​the polygonal area;

[0030] The corner point coverage is calculated, and the area of ​​the calculated polygonal region is divided by the resolution of the calibration image to obtain the corner point coverage.

[0031] In order to solve the above technical problems, another technical solution adopted by the present invention is to provide a calibration device, comprising:

[0032] Center calibration plate; and

[0033] Multiple side marking plates are configured so that each side marking plate can adjust the unfolding angle and its own size.

[0034] The calibration method and calibration device of the present invention have at least the following beneficial effects: the present invention can effectively solve the problem of insufficient corner point coverage caused by the field of view angle limitation of the camera module in the traditional calibration method by designing a calibration device including a central calibration plate and a plurality of measuring calibration plates with adjustable expansion angles and sizes. For camera modules with different field of view angles such as ADAS and AVM, the flexible adjustment of the side calibration plates can improve the corner point coverage rate in the calibration image, thereby significantly improving the accuracy of the intrinsic reference calibration; by adjusting the expansion angle and size of the side calibration plates, the camera modules with different field of view angles can be quickly adapted without frequent replacement of calibration plates or adjustment of calibration. environment, thereby significantly shortening the calibration time and improving the calibration efficiency; compared with the traditional two-dimensional calibration plate which needs to shoot multiple calibration images to improve the accuracy, when the calibration device is adjusted to a three-dimensional calibration light box, only one calibration image needs to be shot, thereby reducing the complexity of operation and improving the calibration efficiency; by integrating the functions of providing illumination and calibration together, displaying the chart used for calibration on the light source board, not only can space and cost be saved, but also uniform and stable illumination can be provided for the chart; by adopting an adaptive corner point algorithm to process the calibration image, the corner point coverage rate can be effectively improved, thereby improving the accuracy of the intrinsic parameter calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0036] Figure 1 It is a schematic structural diagram of an implementation manner of a calibration device of the present invention.

[0037] Figure 2 The present invention is a flow chart of an implementation method of the calibration method.

[0038] Figure 3 Schematic diagram of the flat calibration plate.

[0039] Figure 4 Schematic diagram of a calibration light box with an opening at the bottom.

[0040] Figure 5 A flow chart for adjusting the calibration device to a calibration light box with an open bottom and adjusting the degree of opening thereof according to a specific field of view angle to adapt to the field of view angle.

[0041] Figure 6 for Figure 5 Flowchart of step S23 in FIG.

[0042] Figure 7 It is a calibration image taken before de-distortion when the calibration device is used to perform internal parameter calibration.

[0043] Figure 8 for Figure 7 The calibration image after dedistortion.

[0044] Fig. 9 It is a corner point coverage map when the calibration device is used to perform internal parameter calibration.

[0045] Fig.10 This is a diagram of internal parameter calibration data when the calibration device is used to perform internal parameter calibration.

[0046] Fig.11 for Figure 2 Flow chart of step S3 in FIG.

[0047] Fig.12 This is the internal parameter calibration data diagram when using the corner point algorithm that comes with OpenCV.

[0048] Fig.13 This is the corner coverage map when using OpenCV's built-in corner algorithm.

[0049] Fig.14 This is the internal parameter calibration data diagram when using the adaptive corner point algorithm.

[0050] Fig.15 This is the corner coverage map when the adaptive corner algorithm is used. DETAILED DESCRIPTION

[0051] The present invention will be further described below in conjunction with the accompanying drawings.

[0052] See also Figure 1The calibration device of the present invention includes a central calibration plate and a plurality of side calibration plates arranged around the central calibration plate; the central calibration plate is configured as a rectangular central calibration plate, and the plurality of side calibration plates are configured as four side calibration plates respectively arranged on the four sides of the central calibration plate. Different camera modules have different field of view angles. In order to be applicable to camera modules with various field of view angles, the angle between each of the four side calibration plates and the central calibration plate can be adjusted to avoid the calibration area from being blocked and improve the corner point coverage rate; in order to be applicable to camera modules with various field of view angles, the size of each side calibration plate can be adjusted to provide the largest calibration area possible and improve the corner point coverage rate.

[0053] The central calibration plate is defined as a single light source plate; each side calibration plate includes a mother light source plate and at least one sub-light source plate, the at least one sub-light source plate can be stored in the mother light source plate, and the at least one sub-light source plate can extend out of the mother light source plate from at least one side of the mother light source plate, and the size of the side calibration plate is adjusted by controlling the area of ​​the sub-light source plate extending out of the mother light source plate. In this embodiment, the number of sub-light source plates included in each side calibration plate is limited to two, both sub-light source plates can be stored in the mother light source plate, and the two sub-light source plates can extend out of the mother light source plate from opposite sides of the mother light source plate; the angle between the mother light source plate and the single light source plate is adjusted by controlling the handle of the mother light source plate, and the position of the sub-light source plate extending out or stored in the mother light source plate is adjusted by controlling the handle of the sub-light source plate. In order to save space and cost, and to provide uniform and stable illumination for the chart, this solution integrates the functions of illumination and calibration. The chart used for calibration can be displayed on the light source board, and the chart is a black and white chessboard; and the chart specifications on all light source boards are the same to ensure the consistency and accuracy of calibration. The vertex where the black and white grids intersect in the chart is the corner point. The light source board includes a mother light source board, a sub-light source board, and a single light source board. In this embodiment, all light source boards are the same size.

[0054] See also Figure 2 , is a flow chart of an implementation method of the calibration method of the present invention. This implementation method specifically includes the following steps:

[0055] S1. Provide a calibration device for internal reference calibration.

[0056] The calibration device includes a central calibration plate and a plurality of side calibration plates arranged around the central calibration plate. The calibration device provided in this embodiment is the calibration device described in the above embodiment, and the specific structure is not described here one by one. In specific implementation, the picture card on each light source board in the calibration device can be adjusted according to the size of the field of view, and the number of corner points, the number of chessboard grids, and the size of the grids can be switched. After adjustment, it is imported and displayed on the light source board, and can be stored in the memory of the light source board, and can be directly called when needed.

[0057] S2, side calibration plate for adjusting the field of view angle based on the camera module.

[0058] The expansion angle and / or size of the multiple side calibration plates are adjusted based on the field of view of the camera module. Specifically, if the field of view of the camera module is less than or equal to a first threshold, the side calibration plates of the calibration device are adjusted to expand into a flat calibration plate. In this embodiment, the first threshold is set to 120 degrees. For specific implementation, please refer to Figure 3 , adjust the angle of the mother light source board 1 of each side calibration board so that it is in the same horizontal plane as the single light source board 3, and extend a sub-light source board 2 from each mother light source board 1, and all light source boards are spliced ​​into a flat calibration board. If the field of view of the camera module is greater than the first threshold and less than or equal to the second threshold, the calibration device is adjusted to a calibration light box with an opening at the bottom. In this embodiment, the second threshold is set to 180 degrees. For specific implementation, please refer to Figure 4 , control all sub-light source panels to be retracted into the main light source panel, adjust the angle of each main light source panel to make it perpendicular to the single light source panel, and form a calibration light box with an opening at the bottom, Figure 4 The calibration light box is an inverted bottom-open calibration light box. If the field of view of the camera module is greater than the second threshold, the calibration device is adjusted to a bottom-open calibration light box and its degree of opening is adjusted according to the specific field of view to adapt to the field of view. For specific implementation, please refer to Figure 5 The step of adjusting the calibration device to a calibration light box with an open bottom and adjusting the degree of opening thereof according to a specific field of view angle to adapt to the field of view angle specifically includes the following sub-steps:

[0059] S21. Adjust the light source board.

[0060] Specifically, the included angle between each mother light source panel and the single light source panel is randomly adjusted to be the same and greater than 90 degrees, and the sub-light source panels are extended so that the sub-light source panels extended from adjacent mother light source panels are abutted, that is, the sub-light source panels are extended to the maximum extent to provide the largest calibration area, thereby improving the corner point coverage.

[0061] S22: Capture a calibration image.

[0062] Specifically, the camera module is placed at the center of the calibration device, the calibration device is placed above the camera module, and the calibration device at this time is photographed to obtain a calibration image. In this embodiment, it is not necessary to take multiple calibration images, and only one calibration image is sufficient, because the calibration device is a three-dimensional calibration light box, which can completely cover the field of view of the camera module, especially the edge part, so as to ensure that the edge corners can also be clearly photographed.

[0063] S23, using an adaptive corner point algorithm to process the calibration image to obtain corner point coverage.

[0064] See also Figure 6 , this step S23 includes the following sub-steps:

[0065] S231. Remove background interference.

[0066] Specifically, the RGB color channel information of the calibration image is extracted, and the Alpha channel is ignored to remove the interference of the background and facilitate the observation of the chart.

[0067] S232, corner point detection.

[0068] Specifically, the corner points on the calibration image after the interference is removed are first detected. When implemented specifically, a corner detection algorithm such as Harris corner detection and Shi-Tomas i corner detection can be used.

[0069] S233. Calculate the area.

[0070] Specifically, the detected outermost corner points are connected to form a polygonal area, and the area of ​​the polygonal area is calculated. The polygonal area represents the coverage of the valid corner points on the calibration plate.

[0071] S234, calculate the corner point coverage.

[0072] Specifically, the calculated area of ​​the polygonal region is divided by the resolution of the calibration image to obtain the corner point coverage rate, which reflects the density of valid corner points in the calibration image. The higher the corner point coverage rate, the fewer corner points are not involved in the calculation of the intrinsic parameter calibration, and the higher the accuracy of the intrinsic parameter calibration.

[0073] S24, determining the corner point coverage and adjusting the calibration device.

[0074] Specifically, it is determined whether the corner point coverage is greater than or equal to a set third threshold. If the corner point coverage is greater than or equal to the set third threshold, the calibration device is no longer adjusted, and the quality of the calibrated image has met the requirements; if the corner point coverage is less than the set third threshold, steps S21 to S23 are repeated, and the calibration device is continuously adjusted until the corner point coverage is greater than or equal to the set third threshold. The third threshold can be set according to actual needs. In this embodiment, the third threshold is set to 90%.

[0075] S3. Capture a calibration image for calibration.

[0076] The camera module captures the adjusted calibration image of the calibration device for calibration. Specifically, when the field of view of the camera module is less than or equal to 120 degrees, the calibration device is adjusted to a two-dimensional plane calibration plate, and multiple calibration images need to be taken for calibration; when the field of view of the camera module is greater than 120 degrees, the calibration device is adjusted to a three-dimensional calibration light box, and only one calibration image needs to be taken. Because the three-dimensional calibration light box adapted to the field of view can completely cover the field of view of the camera module, especially the edge part, it can ensure that the edge corners can also be clearly captured, and there is no need to adjust the position of the camera module and take multiple shots.

[0077] Take the AVM camera module with a resolution of 1920*1300, a field of view of HFOV196.3°, and VFOV139.2° as an example (adjust the calibration device according to HFOV), and use the calibration device described in this embodiment for calibration. Figures 7 to 10 It can be seen that the de-distortion effect of the internal reference calibration is normal, and the internal reference data meets the specification requirements. The corner point coverage rate of the internal reference calibration of the AVM camera module can be increased from about 60% to more than 85%.

[0078] See also Fig.11 , this step S3 includes the following sub-steps:

[0079] S31, using an adaptive corner point algorithm to process the calibration image.

[0080] Specifically, the step of using the adaptive corner point algorithm to process the calibration image in this step is the same as the content of step S23. The specific content can be found in step S23, which will not be described in detail here.

[0081] The corner point algorithm that comes with OpenCV can only recognize complete and unobstructed checkerboards. As long as there are corner points that are blocked, the intrinsic parameters will not be calculated. Therefore, when using the pinhole model to calibrate a camera module with a larger field of view, the four corners of the image will have areas that are not covered by the checkerboard corner points due to distortion. When the area that is not covered by the corner points is relatively large, it will have a great impact on the improvement of the corner point coverage rate and the de-distortion effect of the intrinsic parameter calibration, resulting in poor de-distortion effect at the four corners and low accuracy of the internal parameter calibration at the four corners. The adaptive corner point algorithm in this embodiment can recognize checkerboards that partially exceed the boundaries, which allows corner points to cover the corners of the image. When the camera module image is out of focus, the corners of the image will be covered by corner points. Figure 4 When the corner is partially blocked by checkerboard corner points, the unblocked part of the checkerboard corner points can participate in the calculation of the intrinsic parameter calibration, which effectively improves the corner point coverage rate. At the same time, the accuracy of the intrinsic parameter calibration at the corners is improved, and the dedistortion effect will be better. The intrinsic parameter will not be calculated due to the occlusion of some corner points.

[0082] S32, performing internal parameter calibration based on the corner point data obtained after processing using the adaptive corner point algorithm.

[0083] Take the ADAS camera module with a resolution of 1920*1080, a field of view of HFOV95, and VFOV47° as an example. When using the corner point algorithm provided by OpenCV, please refer to Fig.12 and Fig.13 According to the intrinsic calibration data, the corner point coverage is about 85%. There is about 15% area at the edge of the camera module where no corner points are captured for calculation of the intrinsic calibration. This area is equivalent to no intrinsic calibration. No distortion is performed on this area in the captured image, which affects the intrinsic calibration effect. For the same camera module, when the adaptive corner point algorithm described in this embodiment is used, please refer to Fig.14 and Fig.15 It can be seen that the effect of intrinsic calibration de-distortion is normal, and the intrinsic parameter data meets the requirements. The corner point coverage rate of the ADAS module intrinsic calibration is increased from about 85% of the OpenCV built-in corner point algorithm to more than 90% of the adaptive corner point algorithm.

[0084] The present invention can effectively solve the problem of insufficient corner point coverage caused by the limitation of the field of view of the camera module in the traditional calibration method by designing a calibration device including a central calibration plate and a plurality of measuring calibration plates with adjustable expansion angles and sizes. For camera modules with different field of view angles such as ADAS and AVM, the flexible adjustment of the side calibration plates can improve the corner point coverage rate in the calibration image, thereby significantly improving the accuracy of the intrinsic reference calibration; by adjusting the expansion angle and size of the side calibration plates, the camera modules with different field of view angles can be quickly adapted, without the need to frequently replace the calibration plates or adjust the calibration environment, thereby significantly shortening the calibration time. time, and improves the calibration efficiency; compared with the traditional two-dimensional calibration plate which needs to shoot multiple calibration images to improve the accuracy, when the calibration device is adjusted to a three-dimensional calibration light box, only one calibration image needs to be shot, thereby reducing the operation complexity and improving the calibration efficiency; by integrating the functions of providing illumination and calibration together, and displaying the chart used for calibration on the light source board, not only can space and cost be saved, but also uniform and stable illumination can be provided for the chart; by adopting an adaptive corner point algorithm to process the calibration image, the corner point coverage rate can be effectively improved, thereby improving the accuracy of the intrinsic parameter calibration.

Claims

1. A calibration method, characterized in that: The following steps are involved: A calibration device for internal reference calibration is provided, wherein the calibration device comprises a central calibration plate and a plurality of side calibration plates arranged around the central calibration plate, wherein each of the plurality of side calibration plates can adjust an expansion angle and its own size; Adjusting the deployment angles and / or sizes of the plurality of side calibration plates based on the field of view of the camera module; The camera module is enabled to capture the adjusted calibration image of the calibration device for calibration.

2. The calibration method according to claim 1, characterized in that: The central calibration plate is configured as a rectangular central calibration plate, and the multiple side calibration plates are configured as four side calibration plates respectively arranged on the four sides of the central calibration plate; among the four side calibration plates, the angle between each side calibration plate and the central calibration plate can be adjusted, and the size of each side calibration plate itself can be adjusted.

3. The calibration method according to claim 1, characterized in that: Among the multiple side calibration plates, each side calibration plate includes a mother light source plate and at least one sub-light source plate, the at least one sub-light source plate can be received in the mother light source plate, and the at least one sub-light source plate can extend out of the mother light source plate from at least one side of the mother light source plate; the central calibration plate is defined as a single light source plate.

4. The calibration method according to claim 3, characterized in that: The number of sub-light source boards included in each side marking board is limited to two, and both sub-light source boards can be received in the mother light source board, and the two sub-light source boards can extend out of the mother light source board from opposite sides of the mother light source board; All light source boards have the same size, and all light source boards are displayed with a chart for calibration and the specifications of the chart on all light source boards are the same. The light source boards include a mother light source board, a sub-light source board, and a single light source board.

5. The calibration method according to claim 4, characterized in that: The step of adjusting the calibration device according to the field of view of the camera module specifically includes: If the field of view angle of the camera module is less than or equal to the first threshold, the side calibration plates of the calibration device are adjusted to unfold into a flat calibration plate; if the field of view angle of the camera module is greater than the first threshold and less than or equal to the second threshold, the calibration device is adjusted to a calibration light box with an opening at the bottom; if the field of view angle of the camera module is greater than the second threshold, the calibration device is adjusted to a calibration light box with an open bottom and the degree of opening is adjusted according to the specific field of view to adapt to the field of view angle.

6. The calibration method according to claim 5, characterized in that: The step of adjusting the side calibration plates of the calibration device to unfold them into a flat calibration plate specifically includes: adjusting the angle of the mother light source board of each side calibration plate so that it is in the same horizontal plane as the single light source board, and extending a sub-light source board from each mother light source board, and all the light source boards are spliced ​​into a flat calibration plate. The step of adjusting the calibration device to a calibration light box with a bottom opening specifically includes: controlling all sub-light source panels to be retracted into the main light source panel, adjusting the angle of each main light source panel to make it perpendicular to the single light source panel, and forming a calibration light box with a bottom opening.

7. The calibration method according to claim 5, characterized in that: The step of adjusting the calibration device to a calibration light box with an open bottom and adjusting the degree of opening thereof according to a specific field of view angle to adapt to the field of view angle specifically includes the following sub-steps: Randomly adjust the included angle between each mother light source panel and the single light source panel to be the same and greater than 90 degrees, and extend the sub-light source panels so that the sub-light source panels extended from adjacent mother light source panels abut against each other; Placing the camera module at the center of the calibration device, photographing the calibration device at this time, and obtaining a calibration image; Processing the calibration image using an adaptive corner point algorithm to obtain a corner point coverage rate, and determining whether the corner point coverage rate is greater than or equal to a set third threshold; If the corner point coverage is greater than or equal to a set third threshold, the calibration device is no longer adjusted; If the corner point coverage is less than the set third threshold, repeat the above steps and continue to adjust the calibration device until the corner point coverage is greater than or equal to the set third threshold.

8. The calibration method according to claim 1, characterized in that: The step of enabling the camera module to capture the adjusted calibration image of the calibration device for calibration specifically includes the following sub-steps: Using an adaptive corner point algorithm to process the calibration image; Intrinsic parameter calibration is performed based on the corner point data obtained after processing using the adaptive corner point algorithm.

9. The calibration method according to claim 7 or claim 8, characterized in that: The step of processing the calibration image using the adaptive corner point algorithm specifically includes the following sub-steps: Remove background interference, extract the RGB color channel information of the calibration image, and ignore the Alpha channel; Corner detection: detect the corner points on the calibration image after removing interference; Calculate the area by connecting the outermost corner points detected to form a polygonal area, and calculate the area of ​​the polygonal area; The corner point coverage is calculated, and the area of ​​the calculated polygonal region is divided by the resolution of the calibration image to obtain the corner point coverage.

10. A calibration device, characterized in that: include: Center calibration plate; as well as Multiple side marking plates are configured so that each side marking plate can adjust the unfolding angle and its own size.

Images