Camera calibration method and device, equipment and storage medium

Abstract

The invention discloses a camera calibration method and device, equipment and a storage medium. The method comprises the following steps: acquiring a first image acquired by shooting a calibration plate by a camera; wherein the calibration plate comprises a plurality of angular points; performing inverse perspective transformation on the first image based on the internal reference and the initial external reference of the camera to obtain a second image; detecting angular points in the second image to obtain a first angular point detection result; based on the first corner detection result, the first image and the internal reference, solving to obtain a target external reference of the camera; and calculating an error parameter based on the target external parameter, the internal parameter and the first image, and when the error parameter satisfies a preset condition, determining that camera calibration is completed. According to the technical scheme of the invention, the inverse perspective transformation is carried out on the first image shot and collected by the calibration plate, so that the influence of partial camera distortion can be eliminated, the accuracy of angular point detection is improved, and the success rate and accuracy of camera calibration can be improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of intelligent driving, and particularly relates to a camera calibration method and device, equipment and a storage medium. BACKGROUND

[0002] In the intelligent driving technology, a plurality of cameras are usually installed at different positions of a vehicle, and the visual perception data collected by the plurality of cameras can improve the safety and convenience of intelligent driving. Before the vehicle is shipped, the cameras usually need to be calibrated offline (End-of-Line Calibration, EOL) to determine whether the internal and external parameters of the cameras meet the requirements.

[0003] At present, the offline calibration method of the camera mainly arranges a checkerboard target around the vehicle, and calibrates offline the image collected by the camera on the checkerboard target. However, this scheme is affected by camera distortion, vehicle angle, environmental light, etc., resulting in low accuracy of checkerboard corner point detection, and even the checkerboard corner point detection may fail, affecting the accuracy or success rate of the calibration result. SUMMARY

[0004] Generally, the camera calibration method in the related art is directly based on the collected checkerboard image for calibration. This calibration method is affected by camera distortion, vehicle angle, environmental light, etc., resulting in low accuracy of checkerboard corner point detection or even detection failure, affecting the accuracy or success rate of the calibration result. In order to solve the above technical problems, the present disclosure provides a camera calibration method and device, equipment and a storage medium.

[0005] The first aspect embodiment of the present disclosure provides a camera calibration method, comprising: acquiring a first image collected by a camera shooting a calibration board; wherein the calibration board comprises a plurality of corner points; performing inverse perspective transformation on the first image based on the internal parameter and the initial external parameter of the camera to obtain a second image; detecting the corner points in the second image to obtain a first corner point detection result; solving the target external parameter of the camera based on the first corner point detection result, the first image and the internal parameter; calculating an error parameter based on the target external parameter, the internal parameter and the first image, and determining that the camera calibration is completed when the error parameter meets a preset condition.

[0006] In a second aspect, embodiments of the present disclosure provide a camera calibration device, comprising: an acquisition module configured to acquire a first image captured by a camera shooting a calibration board; wherein the calibration board comprises a plurality of corner points; a determination module configured to perform inverse perspective transformation on the first image acquired by the acquisition module based on intrinsic parameters and initial extrinsic parameters of the camera to obtain a second image; a detection module configured to detect the corner points in the second image obtained by the determination module to obtain a first corner point detection result; a processing module configured to solve target extrinsic parameters of the camera based on the first image, the intrinsic parameters, and the first corner point detection result determined by the detection module; and a calibration module configured to calculate error parameters based on the target extrinsic parameters determined by the processing module, the intrinsic parameters, and the first image, and determine that the camera calibration is completed when the error parameters meet a preset condition.

[0007] In a third aspect, embodiments of the present disclosure provide a computer readable storage medium storing a computer program, the computer program being configured to execute the camera calibration method of the first aspect.

[0008] In a fourth aspect, embodiments of the present disclosure provide an electronic device, comprising: a processor; a memory configured to store processor-executable instructions; and the processor configured to read the executable instructions from the memory and execute the instructions to implement the camera calibration method of the first aspect.

[0009] In a fifth aspect, embodiments of the present disclosure provide a computer program product, when instructions in the computer program product are executed by a processor, the camera calibration method of the first aspect is executed.

[0010] Based on the camera calibration method provided by the present disclosure, by performing inverse perspective transformation on the first image captured by shooting the calibration board, and performing corner point detection on the second image obtained by inverse perspective transformation, and then obtaining the camera extrinsic parameters based on the detected corner points, the first image and the camera intrinsic parameters, the success rate of corner point extraction can be improved, and thus the success rate and accuracy of camera calibration can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0011] Figure 1 A flowchart of the camera calibration method provided by an exemplary embodiment of the present disclosure.

[0012] Figure 2A A scene diagram applicable to the camera calibration method provided by an exemplary embodiment of the present disclosure.

[0013] Figure 2B A diagram of a checkerboard provided by an exemplary embodiment of the present disclosure.

[0014] Figure 2C A schematic diagram of a first image and a second image provided for an exemplary embodiment of the present disclosure.

[0015] Figure 3 A flowchart of a method for solving a target extrinsic parameter provided for an exemplary embodiment of the present disclosure.

[0016] Figure 4A A flowchart of a method for solving a target extrinsic parameter provided for another exemplary embodiment of the present disclosure.

[0017] Figure 4B A schematic diagram of a third image provided for an exemplary embodiment of the present disclosure.

[0018] Figure 5A A flowchart of a method for solving a target extrinsic parameter provided for yet another exemplary embodiment of the present disclosure.

[0019] Figure 5B A schematic diagram of a fourth image provided for an exemplary embodiment of the present disclosure.

[0020] Figure 6 A flowchart of a method for determining a first corner point detection result provided for an exemplary embodiment of the present disclosure.

[0021] Figure 7 A flowchart of a camera calibration method provided for another exemplary embodiment of the present disclosure.

[0022] Figure 8 A flowchart of a camera calibration method provided for yet another exemplary embodiment of the present disclosure.

[0023] Figure 9 A structural schematic diagram of a camera calibration apparatus provided for an exemplary embodiment of the present disclosure.

[0024] Figure 10 A structural schematic diagram of a camera calibration apparatus provided for another exemplary embodiment of the present disclosure.

[0025] Figure 11 A structural schematic diagram of an electronic device provided for an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] In order to explain the present disclosure, the example embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings, obviously, the described embodiments are only part of the embodiments of the present disclosure, not all the embodiments, it should be understood that the present disclosure is not limited by the example embodiments.

[0027] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0028] Application Overview

[0029] First, the application scenarios of the present disclosure are introduced. The camera calibration method provided by the embodiments of the present disclosure can be applied to, for example, intelligent driving scenarios, robot automation control scenarios in the industrial field, and any other feasible scenarios.

[0030] In intelligent driving scenarios, vehicles are equipped with surround-view cameras. These cameras provide drivers with a 360-degree panoramic view of the vehicle's surroundings, enhancing the safety and convenience of intelligent driving. Before a vehicle leaves the factory, the surround-view camera must be calibrated offline to determine its external parameters relative to the vehicle and to ensure that the camera's installation angle, among other parameters, meets requirements.

[0031] In related art, offline calibration of surround-view cameras can be performed by placing multiple checkerboard grids (i.e., checkerboard targets) on the ground around the vehicle and capturing a surround-view image from the checkerboard grid. The camera then directly extracts the checkerboard grid corners from the surround-view image and combines them with the previously measured 3D spatial coordinates to determine the extrinsic parameters of the surround-view camera relative to the vehicle. However, due to the large distortion of surround-view cameras and the influence of factors such as the vehicle angle, the camera's intrinsic parameters, and ambient lighting, the related art of directly detecting checkerboard grid corners in the captured raw image is prone to failure, which in turn leads to the failure of offline calibration of the surround-view camera.

[0032] To solve the above problems, an embodiment of the present disclosure provides a camera calibration method. The method performs an inverse perspective transformation on a first image captured by a calibration plate, performs corner point detection on a second image obtained by the inverse perspective transformation, and then obtains camera extrinsic parameters based on the detected corner points, the first image, and the camera intrinsic parameters. In this way, since the influence of some camera distortion on the image can be eliminated in the process of converting the captured first image into a bird's-eye view, the projection of the calibration plate on the two-dimensional plane is closer to the actual layout, thereby improving the success rate of corner point extraction, and thus improving the success rate and accuracy of camera calibration.

[0033] Exemplary Methods

[0034] Figure 1 This is a flow chart of a camera calibration method provided by an exemplary embodiment of the present disclosure. This embodiment can be applied to electronic devices, such as Figure 1 As shown, the method includes the following steps S101 to S104.

[0035] In step S101, a first image captured by the camera on the calibration board is obtained. The calibration board includes a plurality of corner points.

[0036] Exemplarily, the camera in the embodiments of the present disclosure includes, but is not limited to, a surround-view camera, a front-view camera, a rear-view camera, a side-view camera, etc. The embodiments of the present disclosure take the surround-view camera as an example for illustration. In some examples, the calibration board can be a checkerboard composed of black and white grids, or can be other patterns with a plurality of corner points that can be used for camera calibration, and the embodiments of the present disclosure do not limit this.

[0037] Figure 2A A scene diagram to which the camera calibration method provided by an exemplary embodiment of the present disclosure is applicable is shown in FIG. 2. Figure 2A As shown in FIG. 2, taking the surround-view cameras 21 on the vehicle 20 including front-view fisheye cameras, rear-view fisheye cameras, left-view fisheye cameras and rear-view fisheye cameras as an example, when the surround-view cameras 21 on the vehicle 20 are calibrated offline, four calibration boards 22 (such as checkerboards) can be arranged around the vehicle 20, and a first image captured by the surround-view cameras 21 on the calibration boards 22 is obtained. The first image can be an image captured by any one of the fisheye cameras. Figure 2A

[0038] Figure 2B A diagram of the checkerboard provided by an exemplary embodiment of the present disclosure is shown in FIG. 3. Figure 2B As shown in FIG. 3, taking the calibration board 22 as the checkerboard 23 as an example, the checkerboard 23 is composed of black and white grids, and the size of the grid is usually required to be uniform. The checkerboard includes a plurality of corner points 24, and the corner points 24 are intersection points formed at the junctions of the black and white grids in the checkerboard pattern. Figure 2B Only part of the corner points 24 are identified as examples in FIG. 3. By capturing an image of the checkerboard 23 and detecting a plurality of corner points 24 using a corresponding algorithm, the calibration of the camera can be achieved.

[0039] In some examples, after obtaining the first image captured by the camera on the calibration board, the first image captured can be detected first to determine whether the first image meets the requirements. In the case where the first image meets the requirements, the subsequent step S102 is continued. For example, it can be determined whether the camera intrinsic parameters have been calibrated. If the camera intrinsic parameters have been calibrated, the subsequent step is continued.

[0040] In step S102, the first image is inversely perspective transformed based on the intrinsic parameters and the initial extrinsic parameters of the camera to obtain a second image.

[0041] ​Exemplarily, in order to determine the mutual relationship between the three-dimensional geometric position of the surface of the spatial object and the corresponding point thereof in the image captured by the camera, it is necessary to establish a geometric model of camera imaging, and the parameters of the geometric model are camera parameters. The camera parameters need to be obtained through experiments and calculations, and the process of solving the parameters is camera calibration. The camera calibration method in the embodiments of the present disclosure involves the intrinsic parameters of the camera and the extrinsic parameters of the camera. The intrinsic parameters of the camera refer to the geometric and optical characteristics of the camera itself, which are related to the hardware structure of the camera and are irrelevant to the position of the camera in space. The intrinsic parameters of the camera include but are not limited to the focal length of the camera, the principal point coordinates of the camera, the distortion coefficients of the camera, etc.

[0042] In some examples, the intrinsic parameters of the camera in the embodiments of the present disclosure generally refer to the calibrated camera intrinsic parameters or the camera intrinsic parameters that have been calibrated and meet the accuracy requirements. It should be noted that the embodiments of the present disclosure do not limit whether the camera intrinsic parameters have been calibrated or whether the calibration accuracy of the camera intrinsic parameters meets the requirements. The initial extrinsic parameters refer to the position and attitude parameters of the camera in the world coordinate system when the camera is installed or initialized. These parameters generally include a rotation matrix (R) and a translation vector (T), which are used to describe the relative position and direction between the camera coordinate system and the world coordinate system. It should be noted that the initial extrinsic parameters of the camera can be obtained by measurement, and the specific acquisition method of the initial extrinsic parameters of the camera is not limited in the embodiments of the present disclosure.

[0043] In some examples, the embodiments of the present disclosure can perform inverse perspective mapping (IPM) on the captured first image based on the camera intrinsic parameters and the initial extrinsic parameters of the camera to obtain a second image. The basic principle of inverse perspective mapping is to convert the first image from the perspective transformation space to the two-dimensional plane space through the inverse perspective transformation matrix to eliminate the perspective distortion. Moreover, the second image after transformation is an overhead view or a top view including the calibration board, and the overhead view or the top view includes a plurality of corner points of the calibration board.

[0044] Figure 2C The schematic diagram of the first image and the second image provided for an exemplary embodiment of the present disclosure is shown in FIG. 1. Figure 2C As shown in FIG. 1, the first image 25 captured by the camera can be obtained by shooting the checkerboard. Then, the IPM inverse perspective transformation is performed on the first image 25 by using the intrinsic parameters and the initial extrinsic parameters of the camera to obtain the second image 26 after inverse perspective transformation. The second image 26 includes a plurality of corner points of the checkerboard.

[0045] In step S103, the corner points in the second image are detected to obtain a first corner point detection result.

[0046] For example, to mitigate the issue of camera offline calibration failures caused by inaccurate checkerboard corner detection results, the present embodiment performs an inverse perspective transformation on the first image captured by the camera to obtain an IPM-transformed second image. Furthermore, multiple corner points are extracted from the IPM-transformed second image to obtain the first corner detection result.

[0047] In some examples, the first corner point detection result may include the number of corner points of multiple corner points on the calibration plate pattern in the second image. The first corner point detection result may also include the corner point coordinates of multiple corner points on the calibration plate pattern in the second image. For example, if the second image includes a checkerboard pattern, the first corner point detection result may include the number of intersections of black and white squares on the checkerboard in the second image, and the first corner point detection result may also include the coordinates of the intersections of black and white squares on the checkerboard in the second image. It should be noted that the specific implementation method for detecting corner points in the second image and obtaining the first corner point detection result in the embodiments of the present disclosure is not limited. For example, the corner points in the second image can be detected by Harris corner detection, Shi-Tomasi corner detection, Fast corner detection, etc., or can be detected by other methods.

[0048] Step S104: Based on the first corner point detection result, the first image and the intrinsic parameters, the target extrinsic parameters of the camera are obtained.

[0049] For example, an optimization problem may be constructed based on the first corner detection result, the first image, and the camera's intrinsic parameters. The optimization problem is then solved to obtain the camera's target extrinsic parameters.

[0050] Step S105 : calculating error parameters based on the target external parameters, the internal parameters, and the first image, and determining that the camera calibration is completed when the error parameters meet a preset condition.

[0051] For example, embodiments of the present disclosure may calculate error parameters based on the camera's target extrinsic parameters, the camera's intrinsic parameters, and the first image captured by the camera. If the error parameters meet preset conditions, camera calibration is determined to be complete. If the error parameters do not meet the preset conditions, the accuracy of the camera's intrinsic parameters may be rechecked, or the camera's installation position and angle may be reset.

[0052] The camera calibration method provided in an embodiment of the present application performs an inverse perspective transformation on a first image captured by a calibration plate, performs corner detection on a second image obtained by the inverse perspective transformation, and then obtains camera extrinsic parameters based on the detected corner points, the first image, and the camera intrinsic parameters. In this way, since the influence of some camera distortion on the image can be eliminated in the process of converting the captured first image into a bird's-eye view, the projection of the calibration plate on the two-dimensional plane is closer to the actual layout, thereby improving the success rate of corner extraction, thereby improving the success rate and accuracy of camera calibration.

[0053] As shown in the above Figure 3 embodiment, based on the above-mentioned Figure 1 embodiment, the step 104 can include the following steps S1041a-S1042a.

[0054] The step S1041a, in response to the first corner point detection result indicating that the corner point detection is successful, reprojects the first image coordinates of the plurality of corner points on the second image to the first image based on the intrinsic parameters and the initial extrinsic parameters, to obtain the second image coordinates of the plurality of corner points.

[0055] Exemplarily, the first corner point detection result includes the first image coordinates of the plurality of corner points on the second image. Further, in the case that the first corner point detection result indicates that the corner point detection is successful, the first image coordinates of the plurality of corner points on the second image can be reprojected to the first image captured and collected based on the intrinsic parameters of the camera and the initial extrinsic parameters of the camera, to obtain the second image coordinates of the plurality of corner points. Wherein, the image coordinate system is a two-dimensional coordinate system for describing the pixel position in the image, which is used to describe the position of the point in the image relative to the imaging plane of the camera. Therefore, the first image coordinates are the 2D (Two Dimensional) coordinates of the plurality of corner points on the second image.

[0056] In some examples, whether the first corner point detection result is successful can be determined in the following way: since the actual size information of the calibration board is known in advance, the detection size information of the calibration board can be determined through the first corner point detection result, and if the actual size information and the detection size information are the same, it is determined that the corner point detection is successful. For example, as shown in the above Figure 2B , the actual size of the calibration board 22 placed in front of the vehicle 20 is 3*4, the first image is captured and collected by using the front-view fisheye camera on the calibration board 22, the first image is processed to obtain the first corner point detection result, and the detection size of the calibration board 22 is determined to be 3*4 through the first corner point detection result. Since the actual size and the detection size of the calibration board 22 are both 3*4, the corner point detection is successful.

[0057] The step S1042a, based on the second image coordinates of the plurality of corner points, the world coordinates of the plurality of corner points on the calibration board, and the intrinsic parameters, solves the target extrinsic parameters.

[0058] Exemplarily, the world coordinate system is usually used to describe the absolute position of an object in a three-dimensional space, and thus the world coordinates of the plurality of corner points on the calibration board 22 can be obtained. Figure 2B Further, the world coordinates of the plurality of corner points on the calibration board 22, the second image coordinates of the plurality of corner points, and the intrinsic parameters of the camera can be used to construct an optimization problem to solve the target extrinsic parameters of the camera.

[0059] The camera calibration method provided in the embodiments of the present application directly reprojects the positions of the plurality of corner points on the second image after IPM transformation to solve the extrinsic parameters of the camera in the case of successful corner point detection. In this way, since the influence of partial camera distortion on the image can be eliminated in the process of converting the collected first image into an aerial view, the success rate of corner point extraction is improved, and thus the extrinsic parameters of the camera can be accurately solved.

[0060] As shown in the above Figure 4A Based on the embodiments shown in the above Figure 1 The step 104 can include the following steps S1041b-S1043b.

[0061] The step S1041b, in response to the first corner point detection result indicating that the corner point detection fails, corrects the second image to obtain a third image.

[0062] Exemplarily, whether the first corner point detection result fails can be determined in the following manner: since the actual size information of the calibration board is known in advance, the detection size information of the calibration board can be determined through the first corner point detection result, and if the actual size information and the detection size information are different, it is determined that the corner point detection fails. For example, as shown in the above Figure 2B The actual size of the calibration board 22 placed in front of the vehicle 20 is 3*4, the first image is collected by shooting the calibration board 22 with the front-view fisheye camera, the second image is obtained by performing IPM inverse perspective IPM transformation on the first image, and the first corner point detection result of the second image is obtained. The detection size of the calibration board 22 is determined to be 5*3 through the first corner point detection result. Since the actual size and the detection size of the calibration board 22 are different, the corner point detection fails.

[0063] In some examples, correcting the second image to obtain the third image can be warp transformation (affine transformation) on the second image to obtain the third image. The warp transformation is an image processing operation, which can be realized by performing trigonometric function transformation on the pixel points in the second image, so as to twist and deform the second image to transform it into the third image. The second image and the third image have different shapes; for example, the second image is a twisted and deformed rectangle, and the third image is a regular rectangle. Of course, the second image can also be perspective transformed to obtain the third image, and the specific implementation manner of the correction is not limited in the embodiments of the present disclosure.

[0064] In some embodiments, since the initial extrinsic parameters of the camera can not be accurate, the second image generated by the IPM inverse perspective transformation does not conform to the standard (e.g., the checkerboard pattern in the second image is not rectangular), resulting in the detected checkerboard size being different from the actual checkerboard size and thus causing the corner point detection to fail. Therefore, it is necessary to correct the second image so as to correct the non-rectangular checkerboard to a rectangular checkerboard. In the embodiments of the present disclosure, in order to correct the second image, a warp transformation can be performed on the second image, and the warp transformation does not need to use the intrinsic parameters of the camera and the initial extrinsic parameters of the camera, but only needs to use the image coordinates of the plurality of corner points on the checkerboard. For example, taking the actual checkerboard size of 6*3 as an example, if the detected checkerboard size is 4*3, it is determined that the corner point detection on the second image fails. Further, the four corner points of the detected checkerboard can be mapped to a rectangle to determine a transformation matrix, and the third image is obtained by performing the warp transformation on the second image using the transformation matrix.

[0065] Figure 4B A schematic diagram of the third image provided by an exemplary embodiment of the present disclosure is shown in FIG. 4B. As shown in FIG. 4B, in response to the first corner point detection result indicating that the corner point detection fails, the second image is corrected by using the warp transformation to obtain a third image 41. As can be seen, the checkerboard pattern in the corrected third image 41 is a regular rectangular pattern, so that the checkerboard 42 in the third image 41 can be successfully extracted. Figure 4B

[0066] Step S1042b, detecting the corner points in the third image to obtain a second corner point detection result.

[0067] Exemplarily, in order to reduce the problem of camera offline calibration failure caused by the inaccurate corner point detection result of the calibration board in the second image after the IPM transformation, the second image is corrected (e.g., warp transformation) to obtain a third image in the embodiments of the present disclosure. Further, the second corner point detection result of the third image can be obtained by performing the corner point detection on the third image after the correction.

[0068] In some examples, the second corner point detection result can include the number of corner points of the plurality of corner points on the calibration board pattern in the third image. The second corner point detection result can also include the coordinates of the plurality of corner points on the calibration board pattern in the third image. For example, the third image includes a checkerboard, and the second corner point detection result can include the number of intersection points of the black and white squares on the checkerboard in the third image. The second corner point detection result can also include the coordinates of the intersection points of the black and white squares on the checkerboard in the third image. It should be noted that the specific implementation manner of detecting the corner points in the third image to obtain the second corner point detection result is not limited in the embodiments of the present disclosure. For example, the corner points in the third image can be detected by using the Harris corner point detection, Shi-Tomasi corner point detection, Fast corner point detection, or other manners.​

[0069] Step S1043b: Based on the second corner point detection result, the first image and the intrinsic parameters, the target extrinsic parameters are obtained.

[0070] For example, an optimization problem may be constructed based on the second corner point detection result, the first image, and the camera's intrinsic parameters. The optimization problem is then solved to obtain the camera's target extrinsic parameters.

[0071] The camera calibration method provided in the embodiment of the present application obtains a third image by correcting the second image after IPM transformation (e.g., warp transformation); then obtains the camera extrinsic parameters based on the corner points detected in the third image, the first image, and the camera intrinsic parameters. In this way, since the correction of the second image can transform the irregular-shaped calibration plate pattern in the second image into a regular-shaped calibration plate pattern in the third image, and this process does not depend on the initial extrinsic parameters of the camera, even if the initial value error of the camera's extrinsic parameters is large, it is still possible to detect multiple corner points of the calibration plate relatively stably, thereby greatly improving the success rate of offline camera calibration.

[0072] like Figure 5A As shown in the above Figure 1 Based on the illustrated embodiment, step S1043b may include the following steps S11 to S13.

[0073] Step S11: In response to the second corner point detection result indicating that the corner point detection is successful, convert the third image to obtain a fourth image.

[0074] For example, if corner point detection is successful on the third image obtained by warping, the positions (image coordinates) of multiple corner points on the third image can be reprojected to solve the camera's target extrinsic parameters. For example, if a third image is obtained by warping the second image using a first perspective transformation matrix, the first perspective transformation matrix can be inverted to obtain a second perspective transformation matrix, and the second perspective transformation matrix can then be used to transform the third image to obtain a fourth image.

[0075] For example, Figure 5B A schematic diagram of a fourth image provided as an exemplary embodiment of the present disclosure, such as Figure 5B As shown, in response to the second corner point detection result indicating that the corner point detection is successful, the third image is transformed to obtain a fourth image 51. Furthermore, the successfully extracted checkerboard corner points in the third image can be reprojected onto a transformed IPM image (i.e., image 51) to obtain third image coordinates of multiple corner points on the fourth image.

[0076] It is understandable that the specific implementation method of whether the second corner point detection result is successful can refer to the implementation method of whether the first corner point detection result is successful, which will not be repeated here.

[0077] Step S12, based on the intrinsic parameters and the initial extrinsic parameters, re-project the third image coordinates of the plurality of corner points on the fourth image to the first image to obtain fourth image coordinates of the plurality of corner points.

[0078] Exemplarily, after the third image is converted to obtain the fourth image in the embodiments of the present disclosure, the third image coordinates of the plurality of corner points on the fourth image can be re-projected to the first image based on the intrinsic parameters of the camera and the initial extrinsic parameters of the camera to obtain the fourth image coordinates of the plurality of corner points.

[0079] In some examples, the inverse perspective matrix is needed when the third image is re-projected to the first image in the embodiments of the present disclosure. The third image is first converted into an IPM image, and then the IPM image is re-projected to the original image (the first image).

[0080] Step S13, based on the fourth image coordinates of the plurality of corner points, the world coordinates of the plurality of corner points on the calibration board and the intrinsic parameters, solving the target extrinsic parameters.

[0081] Exemplarily, the world coordinate system is usually used to describe the absolute position of an object in a three-dimensional space, and then the world coordinates of the plurality of corner points on the calibration board 22 can be obtained. Figure 2B Further, the world coordinates of the plurality of corner points on the calibration board 22, the fourth image coordinates of the plurality of corner points, and the intrinsic parameters of the camera can be used to construct an optimization problem to solve the target extrinsic parameters of the camera.

[0082] The camera calibration method provided by the embodiments of the present disclosure solves the camera extrinsic parameters based on the positions of the plurality of corner points on the third image (the image after warp transformation) in the case of successful detection of the plurality of corner points on the third image. In this way, since the image after warp transformation can change the pattern of the calibration board from an irregular shape to a regular shape (such as from a non-rectangular shape to a rectangular shape), the target extrinsic parameters of the camera can be obtained after re-projecting the detection results of the plurality of corner points on the image after warp transformation.

[0083] As shown in FIG. 1, the camera calibration method provided by the embodiments of the present disclosure can include the following steps S101-S103. Figure 6 Figure 1 Based on the embodiments shown in the above

[0084] Step S1031, detecting the corner points in the second image to obtain the first image coordinates of at least one corner point.

[0085] Exemplarily, the embodiments of the present disclosure can perform corner point detection on the second image after IPM inverse perspective transformation to obtain the first image coordinates of at least one corner point in the second image. The first image coordinates are the coordinates of at least one corner point in the second image in the image coordinate system corresponding to the second image.​

[0086] It should be noted that the specific implementation of detecting the corner points in the second image and obtaining the first image coordinates of the at least one corner point is not limited in the embodiments of the present disclosure. For example, the corner points in the second image can be detected by means of Harris corner detection, Shi-Tomasi corner detection, Fast corner detection, etc., or can be detected by other means.

[0087] In step S1032, the detection size information of the calibration board is determined based on the first image coordinates of the at least one corner point.

[0088] For example, after the first image coordinates of the at least one corner point are determined, the detection size of the calibration board can be determined based on the first image coordinates of the at least one corner point. For example, the detection size of the calibration board can be determined as 5*3 based on the first image coordinates of a group of corner points (such as a plurality of corner points) in the second image.

[0089] In step S1033, the first corner detection result is determined based on the detection size information and the calibration size information of the calibration board.

[0090] For example, if the actual calibration size information of the calibration board is 6*4 and the detection size information of the calibration board is 5*3, the first corner detection result indicates that the corner point detection fails. Similarly, if the actual calibration size information of the calibration board is 6*4 and the detection size information of the calibration board is also 6*4, the first corner detection result indicates that the corner point detection succeeds.

[0091] The camera calibration method provided by the embodiments of the present disclosure determines the detection size of the calibration board through the first image coordinates of the at least one corner point in the second image, and then compares the detection size with the actual size of the calibration board to obtain the first corner detection result. In this way, it can be determined whether the first corner detection result succeeds or fails simply and quickly.

[0092] It can be understood that the specific implementation of detecting the corner points in the third image to obtain the second corner detection result in step S1042b can refer to the implementation of the aforementioned steps S1031-S1033, which will not be described here.

[0093] As shown in FIG. 10, based on the above-mentioned embodiments shown in FIG. 9, step S105 can include steps S1051a-S1053a as follows. Figure 7 Figure 1 In step S1051a, the world coordinates of the plurality of corner points on the calibration board are re-projected to the first image based on the target extrinsic and intrinsic parameters to obtain the fifth image coordinates of the plurality of corner points.

[0094] In step S1051a, the world coordinates of the plurality of corner points on the calibration board are re-projected to the first image based on the target extrinsic and intrinsic parameters to obtain the fifth image coordinates of the plurality of corner points.

[0095] ​Exemplarily, after determining the target extrinsic parameters of the camera, the embodiment of the present disclosure can reproject the world coordinates of multiple corner points on the calibration plate to the originally captured first image based on the target extrinsic parameters and the intrinsic parameters of the camera to obtain the fifth image coordinates of the multiple corner points.

[0096] Step S1052a: Calculate a first reprojection error in the error parameters based on the fifth image coordinates of the multiple corner points and the sixth image coordinates of the multiple corner points on the first image.

[0097] For example, the image coordinates of multiple corner points on the originally captured first image can be determined and used as the sixth image coordinates of the multiple corner points. Furthermore, the fifth image coordinates of the multiple corner points can be compared with the sixth image coordinates, and a first reprojection error in the error parameter can be calculated based on the comparison result. For example, the distance between the fifth image coordinate and the sixth image coordinate can be determined, and the first reprojection error in the error parameter can be determined based on the distance. This distance includes, but is not limited to, Mahalanobis distance, Euclidean distance, Manhattan distance, Chebyshev distance, and Minkowski distance.

[0098] In some examples, the world coordinate points (3D coordinate points) of multiple corner points on the calibration plate can be reprojected onto the originally captured first image using the solved target extrinsic parameters and known intrinsic parameters to obtain the 2D coordinate points of the multiple corner points; wherein the 2D coordinate points are the fifth image coordinates of the multiple corner points. At the same time, the corner points of the calibration plate that have been successfully extracted through two transformations (the first image is first transformed by the IPM transformation and then by the warp transformation) can be reprojected onto a single-transformation IPM image (the second image after the first image is transformed by the IPM transformation), and then the corner points of the calibration plate projected onto the single-transformation IPM image can be reprojected onto the original image (the first image) to obtain the sixth image coordinates of the multiple corner points on the first image. Furthermore, the disclosed embodiments can construct equations to minimize the reprojection error, thereby optimizing the target extrinsic parameters of the camera.

[0099] Step S1053a: When the first re-projection error meets a preset condition, it is determined that the camera calibration is completed.

[0100] For example, if the distance between the fifth image coordinate and the sixth image coordinate is less than a first preset value, it is determined that the first projection error meets the preset condition, and the camera calibration is completed.

[0101] The camera calibration method provided by the disclosed embodiments determines whether camera calibration is complete by reprojecting the world coordinates of multiple corner points on a calibration plate onto a first image, and then comparing the fifth image coordinates of the multiple corner points with the sixth image coordinates of the multiple corner points on the first image. This allows calibration errors to be determined using reprojection, making it simple and quick to determine whether camera calibration is complete.

[0102] like Figure 8As shown in the above Figure 1 Based on the illustrated embodiment, step S105 may include the following steps S1051b to S1056b.

[0103] Step S1051b: performing an inverse perspective transformation on the first image of the first perspective based on the target external parameter and the internal parameter to obtain a fifth image.

[0104] For example, the first image of the first perspective can be subjected to an IPM inverse perspective transformation based on the determined target external parameters and the known internal parameters to obtain a fifth image. The first image of the first perspective can be an image obtained by capturing the calibration plate by any fisheye camera included in the surround view camera. For example, Figure 2A As shown, the first image of the first perspective includes one of the following: a first image captured by the front-view fisheye camera 21 of the checkerboard 22 in front of the vehicle 20, a first image captured by the rear-view fisheye camera 21 of the checkerboard 22 behind the vehicle 20, a first image captured by the right-view fisheye camera 21 of the checkerboard 22 on the right side of the vehicle 20, and a first image captured by the left-view fisheye camera 21 of the checkerboard 22 on the left side of the vehicle 20.

[0105] Step S1052b, performing an inverse perspective transformation on the first image of the second perspective based on the target external parameter and the internal parameter to obtain a sixth image; wherein, the first image of the first perspective and the first image of the second perspective are images taken by different cameras, and there is a common viewing area between the first perspective and the second perspective, and the common viewing area includes at least one common viewing angle point.

[0106] For example, the first image of the first perspective and the first image of the second perspective are images taken by different cameras, and there is a common viewing area between the first perspective and the second perspective; for example, the first image of the first perspective is Figure 2A The first image captured by the middle front-view fisheye camera 21 is the checkerboard 22 in front of the vehicle 20. The first image of the second perspective is Figure 2A The left-view fisheye camera 21 captures a first image of the checkerboard 22 to the left of the vehicle 20. The first image captured by the front-view fisheye camera 21 and the first image captured by the left-view fisheye camera 21 overlap, and the overlapping area includes at least one corner point on the calibration pattern. This overlapping area is called the common view area, and the at least one corner point in the overlapping area is called the common view point.

[0107] It can be understood that the specific implementation method of performing inverse perspective transformation on the first image of the second perspective based on the target external parameter and the internal parameter to obtain the sixth image can refer to the implementation method of the aforementioned step S1051b, which will not be repeated here.

[0108] Step S1053b: convert the seventh image coordinates of the common viewpoint in the fifth image into the target coordinate system to obtain the eighth image coordinates.

[0109] For example, the target coordinate system in the disclosed embodiments may be a BEV (Bird's Eye View) coordinate system. The BEV coordinate system is a technology that projects three-dimensional environmental information onto a two-dimensional plane. Through multi-sensor data fusion and coordinate transformation, it generates a bird's-eye view of the vehicle's surroundings. Therefore, the BEV coordinate system can be used to map image data collected by different cameras to the same coordinate system. Furthermore, the seventh image coordinates of the common viewpoint in the fifth image can be converted to the BEV coordinate system to obtain the eighth image coordinates.

[0110] Step S1054b: convert the ninth image coordinates of the common viewpoint in the sixth image into the target coordinate system to obtain the tenth image coordinates.

[0111] It can be understood that the specific implementation method of converting the ninth image coordinates of the common viewpoint in the sixth image into the target coordinate system to obtain the tenth image coordinates can refer to the implementation method of the aforementioned step S1053b, which will not be repeated here.

[0112] Step S1055b: Calculate the second reprojection error in the error parameter based on the ninth image coordinate and the tenth image coordinate of the common viewing angle.

[0113] For example, the ninth image coordinate and the tenth image coordinate of the common view point may be compared, and the second reprojection error in the error parameter may be calculated based on the comparison result. For example, the spatial distance between the ninth image coordinate and the tenth image coordinate of the common view point may be determined, and the second reprojection error in the error parameter may be determined based on the spatial distance. The spatial distance includes, but is not limited to, Mahalanobis distance, Euclidean distance, Manhattan distance, Chebyshev distance, and Minkowski distance.

[0114] Step S1056b: When the second re-projection error meets a preset condition, it is determined that the camera calibration is completed.

[0115] For example, if the distance between the ninth image coordinate and the tenth image coordinate is less than a second preset value, the first projection error is determined to meet the preset condition, and the camera calibration is completed. The second preset value can be the same as or different from the first preset value, and those skilled in the art can set it according to actual circumstances.

[0116] The camera calibration method provided by the disclosed embodiments performs an IPM transformation on first images from different viewpoints using determined target extrinsic parameters and known intrinsic parameters. The method then determines the spatial distance between common viewpoints in the transformed IPM images within the same target coordinate system, thereby determining whether the camera calibration is complete. This allows the calibration error to be determined by converting the common viewpoints to the same target coordinate system, making it simple and quick to determine whether the camera calibration is complete.

[0117] In some embodiments, step S105, when the error parameters meet the preset conditions, determining that the camera calibration is completed, includes: when the first reprojection error in the error parameters meets the preset conditions, determining that the camera calibration is completed; and / or, when the second reprojection error in the error parameters meets the preset conditions, determining that the camera calibration is completed.

[0118] For example, in the embodiments of the present disclosure, determining that camera calibration is complete may include one of the following: the first reprojection error satisfies a first preset condition, the second reprojection error satisfies a second preset condition, or the first reprojection error satisfies the first preset condition and the second reprojection error satisfies the second preset condition. The first preset condition and the second preset condition may be the same or different, and the embodiments of the present disclosure are not limited thereto.

[0119] Exemplary devices

[0120] Figure 9 A camera calibration device is provided in an embodiment of the present disclosure, such as Figure 9 As shown, the camera calibration device 900 includes an acquisition module 901 , a determination module 902 , a detection module 903 , a processing module 904 and a calibration module 905 .

[0121] An acquisition module 901 is configured to acquire a first image captured by a camera while photographing a calibration plate, wherein the calibration plate includes a plurality of corner points;

[0122] A determination module 902 is configured to perform an inverse perspective transformation on the first image acquired by the acquisition module based on the intrinsic parameters and initial extrinsic parameters of the camera to obtain a second image;

[0123] A detection module 903 is configured to detect corner points in the second image obtained by the determination module to obtain a first corner point detection result;

[0124] A processing module 904 is configured to obtain target extrinsic parameters of the camera based on the first image, the intrinsic parameters, and the first corner point detection result determined by the detection module;

[0125] The calibration module 905 is configured to calculate error parameters based on the target external parameters, internal parameters, and the first image determined by the processing module, and determine that camera calibration is completed when the error parameters meet a preset condition.

[0126] In some embodiments, the processing module 904 is specifically used to, in response to the first corner point detection result indicating that the corner point detection is successful, reproject the first image coordinates of multiple corner points on the second image to the first image based on the intrinsic parameters and the initial extrinsic parameters to obtain the second image coordinates of the multiple corner points; and solve the target extrinsic parameters based on the second image coordinates of the multiple corner points, the world coordinates of the multiple corner points on the calibration plate, and the intrinsic parameters.

[0127] In some embodiments, as Figure 10 As shown, the processing module 904 includes an image correction unit 9041 , a corner point detection unit 9042 and an external parameter solving unit 9043 .

[0128] An image correction unit 9041 is configured to correct the second image to obtain a third image in response to the first corner point detection result indicating that the corner point detection fails;

[0129] A corner point detection unit 9042 is configured to detect corner points in the third image and obtain a second corner point detection result;

[0130] The extrinsic parameter solving unit 9043 is used to solve the target extrinsic parameter based on the second corner point detection result, the first image and the intrinsic parameter.

[0131] In some embodiments, the extrinsic parameter solving unit 9043 is specifically used to convert the third image to obtain a fourth image in response to the second corner point detection result indicating that the corner point detection is successful; based on the intrinsic parameters and the initial extrinsic parameters, the third image coordinates of multiple corner points on the fourth image are reprojected to the first image to obtain the fourth image coordinates of the multiple corner points; based on the fourth image coordinates of the multiple corner points, the world coordinates of the multiple corner points on the calibration plate and the intrinsic parameters, the target extrinsic parameters are solved.

[0132] In some embodiments, the detection module 903 is specifically used to detect corner points in the second image and obtain the first image coordinates of at least one corner point; determine the detection size information of the calibration plate based on the first image coordinates of the at least one corner point; and determine the first corner point detection result based on the detection size information and the calibration size information of the calibration plate.

[0133] In some embodiments, the calibration module 905 is specifically used to reproject the world coordinates of multiple corner points on the calibration plate to the first image based on the target external parameters and internal parameters to obtain the fifth image coordinates of the multiple corner points; based on the fifth image coordinates of the multiple corner points and the sixth image coordinates of the multiple corner points on the first image, calculate the first reprojection error in the error parameter.

[0134] In some embodiments, the calibration module 905 is specifically used to perform an inverse perspective transformation on the first image of the first perspective based on the target external parameters and the internal parameters to obtain a fifth image; perform an inverse perspective transformation on the first image of the second perspective based on the target external parameters and the internal parameters to obtain a sixth image; wherein, the first image of the first perspective and the first image of the second perspective are images taken by different cameras, and there is a common viewing area between the first perspective and the second perspective, and the common viewing area includes at least one common viewing angle point; convert the seventh image coordinates of the common viewing angle point in the fifth image to the target coordinate system to obtain the eighth image coordinates; convert the ninth image coordinates of the common viewing angle point in the sixth image to the target coordinate system to obtain the tenth image coordinates; based on the ninth image coordinates and the tenth image coordinates of the common viewing angle point, calculate the second reprojection error in the error parameter.

[0135] In some embodiments, the calibration module 905 is specifically configured to determine that the camera calibration is completed when a first reprojection error in the error parameters satisfies a preset condition; and / or determine that the camera calibration is completed when a second reprojection error in the error parameters satisfies a preset condition.

[0136] The beneficial technical effects corresponding to the exemplary embodiment of the camera calibration device 900 can be found in the corresponding beneficial technical effects of the exemplary method section above, and will not be repeated here.

[0137] Exemplary electronic devices

[0138] Figure 11 A structural diagram of an electronic device 110 provided in an embodiment of the present disclosure includes at least one processor 111 and a memory 112.

[0139] The processor 111 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 110 to perform desired functions.

[0140] The memory 112 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 111 may execute the one or more computer program instructions to implement the camera calibration method and / or other desired functions of the various embodiments of the present disclosure described above.

[0141] In one example, the electronic device 110 may further include an input device 113 and an output device 114 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0142] The input device 113 may also include, for example, a keyboard, a mouse, etc.

[0143] The output device 114 can output various information to the outside, and may include, for example, a display, a speaker, a printer, a communication network and its connected remote output devices, etc.

[0144] Of course, to simplify, Figure 11 Only some of the components related to the present disclosure in the electronic device 110 are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, the electronic device 110 may further include any other appropriate components according to specific application scenarios.

[0145] Exemplary computer program products and computer-readable storage media

[0146] In addition to the above-mentioned methods and devices, embodiments of the present disclosure may also provide a computer program product, including computer program instructions, which, when executed by a processor, cause the processor to execute the steps in the camera calibration method of various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section.

[0147] The computer program product may be written in any combination of one or more programming languages ​​to implement the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0148] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the processor executes the steps in the camera calibration method of various embodiments of the present disclosure described in the above “Exemplary Method” section.

[0149] The computer readable storage medium can be any combination of one or more computer readable medium(s). The computer readable medium can be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include the following: an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0150] The above describes the basic principles of the present disclosure in combination with specific embodiments, but the advantages, benefits, effects and the like mentioned in the present disclosure are only examples and are not limiting, and it cannot be considered that they are necessarily possessed by each embodiment of the present disclosure. In addition, the above specific details of the disclosure are only for the purpose of example and for the purpose of understanding, and are not limiting, and the above details do not limit the present disclosure to necessarily adopt the above specific details for implementation.

[0151] Those skilled in the art can make various modifications and variations to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A camera calibration method, comprising: Acquire a first image captured by a camera of a calibration plate; wherein the calibration plate includes a plurality of corner points; Performing an inverse perspective transformation on the first image based on the intrinsic parameters and initial extrinsic parameters of the camera to obtain a second image; Detecting corner points in the second image to obtain a first corner point detection result; Obtaining target extrinsic parameters of the camera based on the first corner detection result, the first image, and the intrinsic parameters; An error parameter is calculated based on the target external parameter, the internal parameter, and the first image, and when the error parameter satisfies a preset condition, it is determined that the camera calibration is completed.

2. The method according to claim 1, wherein The step of obtaining a target extrinsic parameter of the camera based on the first corner detection result, the first image, and the intrinsic parameter includes: In response to the first corner detection result indicating that the corner detection is successful, reprojecting the first image coordinates of the multiple corner points on the second image to the first image based on the intrinsic parameters and the initial extrinsic parameters to obtain second image coordinates of the multiple corner points; The target extrinsic parameters are solved based on the second image coordinates of the multiple corner points, the world coordinates of the multiple corner points on the calibration plate, and the intrinsic parameters.

3. The method according to claim 1, wherein The step of obtaining a target extrinsic parameter of the camera based on the first corner detection result, the first image, and the intrinsic parameter includes: In response to the first corner detection result indicating that the corner detection fails, correcting the second image to obtain a third image; Detecting corner points in the third image to obtain a second corner point detection result; The target extrinsic parameter is obtained based on the second corner detection result, the first image, and the intrinsic parameter.

4. The method according to claim 3, wherein: The step of obtaining the target extrinsic parameter based on the second corner detection result, the first image, and the intrinsic parameter includes: In response to the second corner point detection result indicating that the corner point detection is successful, converting the third image to obtain a fourth image; reprojecting the third image coordinates of the multiple corner points on the fourth image to the first image based on the intrinsic parameters and the initial extrinsic parameters to obtain fourth image coordinates of the multiple corner points; The target extrinsic parameters are solved based on the fourth image coordinates of the multiple corner points, the world coordinates of the multiple corner points on the calibration plate, and the intrinsic parameters.

5. The method according to any one of claims 1 to 4, wherein The detecting corner points in the second image to obtain a first corner point detection result includes: detecting corner points in the second image to obtain first image coordinates of at least one corner point; Determining detection size information of the calibration plate based on the first image coordinates of the at least one corner point; A first corner point detection result is determined based on the detected size information and the calibration size information of the calibration plate.

6. The method according to any one of claims 1 to 4, wherein The calculating the error parameter based on the target extrinsic parameter, the intrinsic parameter and the first image includes: reprojecting the world coordinates of a plurality of corner points on the calibration plate onto the first image based on the target extrinsic parameters and the intrinsic parameters to obtain fifth image coordinates of the plurality of corner points; A first reprojection error in the error parameter is calculated based on the fifth image coordinates of the multiple corner points and the sixth image coordinates of the multiple corner points on the first image.

7. The method according to claim 6, wherein: The calculating the error parameter based on the target extrinsic parameter, the intrinsic parameter and the first image includes: Performing an inverse perspective transformation on the first image from the first perspective based on the target extrinsic parameter and the intrinsic parameter to obtain a fifth image; performing an inverse perspective transformation on the first image from the second perspective based on the target extrinsic parameter and the intrinsic parameter to obtain a sixth image; wherein the first image from the first perspective and the first image from the second perspective are images captured by different cameras, and the first perspective and the second perspective have a common viewing area, and the common viewing area includes at least one common viewing angle point; Converting the seventh image coordinates of the common viewpoint in the fifth image into the target coordinate system to obtain the eighth image coordinates; Converting the ninth image coordinate of the common view point in the sixth image to the target coordinate system to obtain a tenth image coordinate; A second reprojection error in the error parameter is calculated based on the ninth image coordinate and the tenth image coordinate of the common viewing angle.

8. The method according to claim 7, wherein: When the error parameter satisfies a preset condition, determining that the camera calibration is completed includes: When the first reprojection error in the error parameters satisfies the preset condition, determining that the camera calibration is completed; and / or, When the second re-projection error in the error parameters satisfies the preset condition, it is determined that the camera calibration is completed.

9. A camera calibration device, comprising: An acquisition module, configured to acquire a first image captured by a camera while photographing a calibration plate; wherein the calibration plate includes a plurality of corner points; a determination module, configured to perform an inverse perspective transformation on the first image acquired by the acquisition module based on the intrinsic parameters and initial extrinsic parameters of the camera to obtain a second image; a detection module, configured to detect corner points in the second image obtained by the determination module to obtain a first corner point detection result; a processing module, configured to obtain a target extrinsic parameter of the camera based on the first image, the intrinsic parameter, and the first corner point detection result determined by the detection module; A calibration module is configured to calculate an error parameter based on the target extrinsic parameter, the intrinsic parameter, and the first image determined by the processing module, and determine that the camera calibration is completed when the error parameter satisfies a preset condition.

10. A computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the camera calibration method according to any one of claims 1 to 8.

11. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the camera calibration method according to any one of claims 1 to 8.

Images