Calibration Method, System, Storage Medium, Program Product, Electronic Device and Vehicle

Abstract

The present application relates to a calibration method, system, storage medium, program product, electronic device and vehicle, and relates to the field of image processing. Among them, the calibration method is applied to a vehicle, and the calibration method mainly includes: using a target method to perform distortion correction on a target projection image, where the target projection image is an image projected onto an irregular curved surface related to the vehicle. Through the above technical solution, a suitable distortion correction method can be selected to perform projector projection distortion correction, effectively solving the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface related to the vehicle, thereby improving the aesthetic degree of the projection screen relative to the driver's field of vision and enhancing the user experience.

Description

Technical Field

[0001] This application relates to the field of image processing, and in particular, to a correction method, system, storage medium, program product, electronic device, and vehicle. Background Art

[0002] In related technologies, the projector distortion correction method usually identifies the imaging parameters during the projection process and the image features of the projection image on the projection plane to determine the correction parameters for distortion correction. The existing distortion correction methods are mostly applied to the projection images on regular planes, and the de-distortion effect is not good when projecting on the irregular curved surfaces related to vehicles. Summary of the Invention

[0003] Embodiments of this application provide a correction method, system, storage medium, program product, electronic device, and vehicle to at least partially solve the above technical problems.

[0004] To achieve the above object, according to the first aspect of this application, a correction method is provided, which is applied to a vehicle. The method includes: performing distortion correction on a target projection image using a target method;

[0005] wherein the target projection image is an image projected on an irregular curved surface related to the vehicle.

[0006] Optionally, the irregular curved surface includes the surface of the A-pillar of the vehicle.

[0007] Optionally, the target method includes one of a perspective transformation method, a pixel matching method, a camera calibration method, a distortion model algorithm, and a multi-view geometry image algorithm.

[0008] Optionally, before performing distortion correction on the target projection image using the target method, the method further includes:

[0009] determining the target method.

[0010] Optionally, the determining the target method includes:

[0011] determining the target method according to a first camera image and a first projection image;

[0012] wherein the first projection image is an image projected on the irregular curved surface;

[0013] the first camera image is an image obtained by the camera photographing the first projection image.

[0014] Optionally, the camera is a human eye measurement camera installed on the vehicle.

[0015] Optionally, determining the target method according to the first camera image and the first projection image includes:

[0016] Generating a target mapping table based on the center point in the first camera image and the corrected point in the first projection image;

[0017] Determining the target method according to the target mapping table.

[0018] Optionally, generating the target mapping table based on the center point in the first camera image and the corrected point in the first projection image includes:

[0019] Obtaining a first coordinate of the center point in the first camera image in the camera image coordinate system;

[0020] Obtaining a second coordinate of the corrected point in the first projection image in the projection image coordinate system;

[0021] Generating the target mapping table between the center point and the corrected point based on the first coordinate and the second coordinate.

[0022] Optionally, before obtaining the first coordinate of the center point in the first camera image in the camera image coordinate system, the method further includes:

[0023] Performing image enhancement on the first camera image to identify the center point.

[0024] Optionally, obtaining the first coordinate of the center point in the first camera image in the camera image coordinate system includes:

[0025] Obtaining a set of detection points based on the first camera image;

[0026] Using a clustering algorithm to determine the first coordinate;

[0027] Wherein, the set of detection points is determined by identifying dot matrix elements in the first camera image;

[0028] The first coordinate is determined by performing clustering analysis on the set of detection points.

[0029] Optionally, obtaining the second coordinate of the corrected point in the first projection image in the projection image coordinate system includes:

[0030] Determining the second coordinate of the corrected point based on the spatial relationship of the corrected point in the first projection image.

[0031] Optionally, generating the target mapping table between the center point and the corrected point based on the first coordinate and the second coordinate includes:

[0032] Generate a first mapping table between the center point and the correction point based on the first mapping relationship between the first coordinate and the second coordinate;

[0033] When the number of the center points is less than or equal to the number of the correction points, determine the repeated mapping points in the projection image coordinate system based on the first mapping table and the second coordinate;

[0034] Determine a second mapping table between the center points to be filled and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points;

[0035] Generate the target mapping table based on the first mapping table and the second mapping table;

[0036] Wherein, the repeated mapping points include at least two correction points having a first mapping relationship with the same center point.

[0037] Optionally, the determining a second mapping table between the center points to be filled and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points includes:

[0038] Scan the first camera image with a convolution kernel based on the first coordinate of the center point to be filled to generate a third coordinate of the center point to be filled in the camera image coordinate system;

[0039] Generate the second mapping table between the center points to be filled and the repeated mapping points based on the second coordinate and the third coordinate.

[0040] Optionally, the determining a second mapping table between the center points to be filled and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points includes:

[0041] Offset the first coordinate of the center point to be filled in a first direction by a predetermined number of pixels based on the first coordinate of the center point to be filled to generate a fourth coordinate of the center point to be filled in the camera image coordinate system;

[0042] Generate the second mapping table between the center points to be filled and the repeated mapping points based on the second coordinate and the fourth coordinate.

[0043] Optionally, the determining the target method according to the target mapping table includes:

[0044] Obtain the regional coordinates of the projection area in the camera image coordinate system according to the target mapping table;

[0045] Processing the regional coordinates by at least two distortion correction methods to obtain at least two corrected camera images;

[0046] Determining the target method according to at least two corrected camera images;

[0047] Wherein, the projection area is the area for displaying the first projection image on the irregular curved surface; the target method is one of at least two of the distortion correction methods.

[0048] Optionally, the processing the regional coordinates by at least two distortion correction methods to obtain at least two corrected camera images includes:

[0049] Based on the target mapping table and the projection area, using at least two of the distortion correction methods to determine at least two corrected projection images, and using the camera to capture at least two of the corrected projection images to obtain at least two of the corrected camera images.

[0050] Optionally, the determining the target method according to at least two corrected camera images includes:

[0051] Determining the coincidence degree of at least two of the corrected camera images;

[0052] Determining the target method according to the value of the coincidence degree; wherein, the coincidence degree is determined according to the coincidence area between at least two of the corrected camera images.

[0053] According to a second aspect of the present application, there is provided a correction system applied to a vehicle, the system comprising:

[0054] An in-vehicle projector for projecting a target projection image on an irregular curved surface related to the vehicle;

[0055] A distortion correction module for performing distortion correction on the target projection image using the above correction method.

[0056] According to a third aspect of the present application, there is provided a non-transitory computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above correction method is implemented.

[0057] According to a fourth aspect of the present application, there is also provided a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the above correction method are implemented.

[0058] According to a fifth aspect of the present application, there is provided an electronic device, including: a memory on which a computer program is stored; a processor for executing the computer program in the memory to implement the above correction method.

[0059] According to a sixth aspect of the present application, there is provided a vehicle including the above-mentioned calibration system or electronic device, or implementing the calibration method described in any of the above embodiments.

[0060] The advantages of the present application are as follows:

[0061] The present application relates to a calibration method, system, storage medium, program product, electronic device and vehicle, and relates to the field of image processing. Among them, the calibration method mainly includes: performing distortion correction on a target projection image using a target method, where the target projection image is an image projected on an irregular curved surface related to the vehicle. Through the above technical solution, a suitable distortion correction method can be selected for projector projection distortion correction, effectively solving the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface related to the vehicle, thereby improving the aesthetic degree of the projection screen relative to the driver's field of view and enhancing the user experience.

[0062] Other features and advantages of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative efforts.

[0064] In order to more fully understand the present application and its beneficial effects, the following will be described in conjunction with the drawings, where the same reference numerals represent the same parts in the following description.

[0065] Figure 1 is a flowchart of the steps of a calibration method provided by an embodiment of the present application;

[0066] Figure 2 is a flowchart of the steps of another calibration method provided by an embodiment of the present application;

[0067] Figure 3a is a top view of a layout diagram of a calibration device provided by an embodiment of the present application;

[0068] Figure 3b is a side view of a layout diagram of a calibration device provided by an embodiment of the present application;

[0069] Figure 4 is a flowchart of the steps of another calibration method provided by an embodiment of the present application;

[0070] Figure 5It is a flowchart of steps of another calibration method provided by an embodiment of the present application;

[0071] Figure 6 It is a schematic architecture diagram of a calibration system provided by an embodiment of the present application;

[0072] Figure 7 It is a schematic architecture diagram of a vehicle provided by an embodiment of the present application.

[0073] Explanation of reference numerals:

[0074] 10, vehicle; 300, calibration system; 301, human eye measurement camera; 302, vehicle cockpit console; 303, in-vehicle projector; 304, A-pillar; 305, distortion correction module. Detailed implementation manners

[0075] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0076] According to the first aspect of the present application, an embodiment of the present application provides a calibration method. Please refer to Figure 1 , the calibration method provided by the embodiment of the present application includes the following steps.

[0077] Step S100: Determine the target method.

[0078] Step S200: Perform distortion correction on the target projection image using the target method. Among them, the target projection image is an image projected on an irregular curved surface related to the vehicle.

[0079] In some embodiments, the irregular curved surface includes the surface of the A-pillar 304 of the vehicle 10.

[0080] In some embodiments, the target method includes one of a perspective transformation method, a pixel matching method, a camera calibration method, a distortion model algorithm, and a multi-view geometry image algorithm.

[0081] It should be understood that the irregular surfaces related to the vehicle 10 include the irregular surfaces inside the vehicle 10 and the irregular surfaces outside the vehicle 10. In the design of the vehicle 10, the cockpit A-pillar 304 refers to the pillars on both sides of the front windshield on the driver's and front passenger's sides. These pillars structurally support the roof and play an important role in the overall safety and stability of the vehicle 10. However, due to their position and shape, the A-pillar 304 may cause a certain degree of obstruction to the driver's field of vision. The irregular surfaces related to the vehicle 10 can be the inner surface of the A-pillar 304 facing the inside of the vehicle 10, the outer surface of the A-pillar 304 facing away from the inside of the vehicle 10, or the welcome light carpet projected on the irregular ground. This application does not make any limitations in this regard.

[0082] Exemplarily, the target projection image is an image projected on the irregular surface of the A-pillar 304 by the in-vehicle projector 303. Specifically, the projection information may include navigation information, warning information, or other augmented reality content, etc. Since the surface of the A-pillar 304 may have a relatively complex geometric shape, it may cause linear or non-linear distortion of the target projection image, resulting in difficulty in recognizing the above projection information in the driver's field of vision. Therefore, it is necessary to select a suitable target method to be applied to the target projection image projected on the cockpit A-pillar 304 to effectively solve the projection distortion problem on the cockpit A-pillar 304 and ensure the accurate transmission of the projection information and the improvement of the driving experience.

[0083] Based on this, in step S100, a suitable target method is determined. The selection of the specific target method may include at least one of the following distortion correction methods:

[0084] Perspective transformation method: Suitable for scenarios that need to correct the distortion caused by the change of perspective.

[0085] Pixel matching method: Corrects the distortion by matching the pixel points in the image and is suitable for fine correction.

[0086] Camera calibration method: Uses the internal and external parameters of the camera for correction and is suitable for scenarios that require high-precision correction.

[0087] Distortion model algorithm: Corrects the distortion based on a specific mathematical model and is suitable for scenarios with known distortion characteristics.

[0088] Multi-view geometry image algorithm: Uses the multi-view geometry principle for correction and is suitable for scenarios that require comprehensive multi-angle information.

[0089] In the embodiment of this application, according to the characteristics of the cockpit A-pillar 304, a suitable target method is selected. In step S200, the selected target method is used to perform distortion correction on the image projected on the irregular surface.

[0090] Through the above technical solution, an appropriate distortion correction method can be selected for the projector projection distortion correction, effectively solving the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface related to the vehicle, thereby improving the aesthetic degree of the projection screen relative to the driver's field of view and enhancing the user experience.

[0091] In some embodiments, the above step S100 includes:

[0092] Step S110: Determine a target method according to the first camera image and the first projection image;

[0093] Wherein, the first projection image is an image projected onto an irregular curved surface;

[0094] The first camera image is an image obtained by the camera photographing the first projection image.

[0095] In some embodiments, the camera is the human eye measurement camera 301 installed on the vehicle.

[0096] The human eye measurement camera 301 is a specially designed device for simulating the position and view of the human eye. In the design of the vehicle 10, the camera is installed at a place simulating the position of the human eye, usually near the head position of the driver's seat, which can ensure that the captured image is similar to the view seen by the human eye in the actual scene. At the same time, the viewing angle of the camera is designed to match the field of view of the human eye, specifically including the horizontal viewing angle and the vertical viewing angle, so as to capture an image similar to the natural field of view of the human eye.

[0097] Using the human eye measurement camera 301 can more accurately capture the image distortion that the driver may see during actual driving, which helps to select a more appropriate target method for distortion correction and ensure that the corrected image is clearly visible in the driver's field of view.

[0098] In the above step S110, the in-vehicle projector 303 is used to project a first projection image onto an irregular curved surface such as the A-pillar 304, the human eye measurement camera 301 is used to photograph the first projection image to obtain a first camera image, and the first projection image and the second camera image are processed and analyzed to select the most appropriate target method for distortion correction, effectively solving the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface and ensuring the accurate transmission of information.

[0099] In some embodiments, the above step S110 specifically includes the following steps:

[0100] Step S111: Generate a target mapping table according to the center point in the first camera image and the correction point in the first projection image;

[0101] Step S112: Determine the target method according to the target mapping table.

[0102] In some embodiments, the above step S111 includes the following steps, as shown in Figure 2 shown:

[0103] Step S1111: Obtain the first coordinate of the center point in the first camera image in the camera image coordinate system;

[0104] Step S1112: Obtain the second coordinate of the calibration point in the first projection image in the projection image coordinate system;

[0105] Step S1113: Generate a target mapping table between the center point and the calibration point based on the first coordinate and the second coordinate.

[0106] In the embodiments of the present application, based on the fixed cockpit layout of the vehicle 10, specifically, the fixed position of the in-vehicle projector 303, the fixed position of the camera (specifically, the human eye measurement camera 301), and the fixed installation position of the A-pillar 304. Among them, the in-vehicle projector 303 can be installed in the center console 302 of the vehicle cockpit. Exemplarily, reference can be made to Figures 3a to 3b shown. It should be understood that this cockpit layout varies according to different vehicle models.

[0107] Each point in the image projected by the in-vehicle projector 303 is a calibration point, and the image formed by these calibration points is the first projection image. Based on the first projection image, a projection image coordinate system is established. Specifically, the upper left corner of the first projection image is used as the origin, the direction from the upper left corner to the upper right corner is the positive x-axis direction, and the direction from the upper left corner to the lower left corner is the positive y-axis direction. The first projection image is a dot matrix image set covering all pixel points, and the composition unit of these dot matrices is a small graphic with a certain regular set shape, including but not limited to circles, rhombuses, and squares.

[0108] It should be understood that since the first projection image is an image projected on an irregular curved surface such as the A-pillar 304 of the cockpit, the calibration points in the first projection image will be distorted. The distorted calibration points are photographed by the camera to obtain the first camera image. Each point in the first camera image is a center point, and a camera image coordinate system is established based on the first camera image. Specifically, the upper left corner of the first camera image is used as the origin, the direction from the upper left corner to the upper right corner is the positive x-axis direction, and the direction from the upper left corner to the lower left corner is the positive y-axis direction.

[0109] As a specific example, taking the first projection image with a resolution of 640×400 as an example, there are a total of 640×400 pixel points. The first projection image is split into multiple dot matrix images, and each dot matrix image contains a X × Y dot matrix, where X is the number of columns, YLet \(n\) be the number of rows, \(c\) be the column interval and \(r\) be the row interval between each point in the dot matrix. Therefore, each row in the first projection image contains calibration points, and each column contains calibration points. It should be understood that when X or Y is 1, c or r is 0 respectively.

[0110] By translating the dot matrix, different dot matrix images can be created. Finally, all dot matrix images are combined to include all pixels in the first projection image.

[0111] Exemplarily, the translation distance of the dot matrix is 1 pixel. Since the moving distance between dot matrix images is small, in order to clearly separate different dot matrix images, a serial number image can be created in the middle of adjacent dot matrix images. This image is used to indicate the serial number of the projection image of the next frame of the projector. At the same time, each dot matrix image is named, and the naming format includes but is not limited to: naming based on the coordinate serial number of the first point. Thus, the coordinates of other points can be determined through the spatial relationship between dot matrix elements.

[0112] As a specific example, set X , Y as numbers greater than 1, that is, multiple rows and multiple columns are collected simultaneously. This method can reduce the number of dot matrix images, thereby reducing the time for collecting data and optimizing the calibration process. However, there are certain requirements for the subsequent algorithm for identifying calibration points. As another specific example, set X as 1 or Y as 1, that is, the dot matrix is a single column or a single row. In this case, the coordinates of each calibration point can be easily identified when identifying calibration points subsequently. However, this method will cause an excessive number of dot matrix images and increase the cost of collecting data.

[0113] It should be understood that the above resolution graphics are only a specific example, and images with other resolutions are still within the protection scope of this application, and this application does not make specific limitations on this.

[0114] In step S1112, obtain the second coordinates of the calibration points in the projection image coordinate system. The set of the second coordinates of all calibration points is , where x is the coordinate value corresponding to the X-axis in the projection image coordinate system, y is the coordinate value corresponding to the Y-axis in the projection image coordinate system, P is the set of calibration points.

[0115] The layout schematic diagram of the calibration device used in the calibration method provided by the embodiments of this application is referred to Figures 3a to 3bAs shown, obviously, the field of view angle of the camera is larger than the projection range. The camera used is the human eye measurement camera 301, simulating the position and view of the driver's human eye, that is, the camera is relatively close to the A-pillar 304. Exemplarily, the distance is dozens of centimeters. Therefore, if the dot matrix unit in the first projection image is a single pixel, in the first camera image obtained by projection and shooting, the corresponding pixels will be very tiny, resulting in difficulties in identifying the subsequent center points. Based on the above problems, in the embodiments of the present application, a circle with a radius of 1, that is, a diameter of 2, is used, and the center of each small circle is a calibration point of the dot matrix. By shooting the set of dot matrix images, a first camera image containing all the pixels of the first projection image can be obtained.

[0116] In step S1111, obtain the first coordinates of the center points in the camera image coordinate system, and the set of the first coordinates of all center points is , where X is the coordinate value corresponding to the X-axis in the camera image coordinate system, Y is the coordinate value corresponding to the Y-axis in the camera image coordinate system, C is the set of identified center points.

[0117] In step S1113, based on the first coordinates of the calibration points obtained in step S1111 and the second coordinates of the center points obtained in step S1112, a target mapping table between the center points and the calibration points can be generated.

[0118] In step S112, determine the target method according to the target mapping table to de-distort the target projection pattern.

[0119] Through the above steps, the mapping relationship between the center points and the calibration points can be established, and the target mapping relationship between the center points and the calibration points can be generated to select an appropriate target method for de-distortion, so as to solve the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface.

[0120] Refer to Figure 2 As shown, in some embodiments, before the above step S1111, the calibration method further includes:

[0121] Step S1110: Perform image enhancement on the first camera image to identify the center points.

[0122] It should be understood that the first camera image is obtained by the camera shooting the above set of dot matrix images containing X × Y dot matrix, and each dot matrix element is a geometric image. Due to various objective factors such as the inconsistent clocks of the camera and the projector and the overly dark image, the clarity of the dot matrix images in the first projection image projected by the projector is insufficient, and further the clarity of the first camera image taken by the camera is also insufficient, which brings difficulties to the subsequent identification of the center points.

[0123] Based on this, before identifying the center point in the first camera image, it is necessary to perform image enhancement on the first camera image to increase the accuracy of subsequent center point identification.

[0124] Specifically, the first camera image obtained by the camera shooting the first projection image undergoes steps such as gamma transformation, image dilation, binarization, and image erosion, so as to transform the first camera image with complex colors into a simple black-and-white dot matrix image and eliminate noise, thereby reducing the difficulty of subsequent center point identification.

[0125] In some embodiments, the above step S1111 includes the following steps:

[0126] Step S301: Obtain a set of detection points based on the first camera image;

[0127] Step S302: Use a clustering algorithm to determine the first coordinate;

[0128] Among them, the set of detection points is determined by identifying the dot matrix elements in the first camera image;

[0129] The first coordinate is determined by performing clustering analysis on the set of detection points.

[0130] The center point detection is based on the idea that the dot matrix elements are regular symmetric geometric figures. By performing edge detection, the edge points of each dot matrix element figure in the first camera image are obtained. Then, by searching for the contour and calculating the centroid, a set of detection points is formed.

[0131] Exemplarily, the Canny operator can be used for edge detection. The Canny operator is a multi-stage edge detection algorithm. Specifically, first, Gaussian filtering is performed on the image to reduce the influence of noise; then, the gradient intensity and direction of the image are calculated, usually using the Sobel operator to calculate the gradient; next, by suppressing non-maximum values, the edges are refined, that is, it is checked whether the pixel is a local maximum in the gradient direction. If not, it is suppressed to zero; then, high and low double thresholds are used to detect strong edges and weak edges. Strong edges are directly retained, and weak edges are retained if they are connected to strong edges, otherwise they are suppressed; finally, by connecting weak edges and strong edges, the final edge map is formed. By the above method, the interference of non-edge points can be reduced. This method is suitable for scenarios that require precise edge detection and can effectively reduce the influence of noise and provide refined edges.

[0132] Exemplarily, the Sobel operator can also be used for edge detection. The Sobel operator is a simple edge detection method mainly used to calculate the gradient of an image. Since the Sobel operator does not use non-maximum suppression and double threshold methods, the edges may be relatively rough and are suitable for rapid detection.

[0133] The embodiments of the present application do not limit the specific edge detection method used. In actual use, it can be selected according to specific application requirements and computing resources.

[0134] After obtaining the detection point set, a clustering algorithm is used to obtain the points of different element graphics, including but not limited to DBSCAN, K-means, etc. The present application does not limit this. Exemplarily, using the DBSCAN algorithm, the minimum number of points is set to 2, and the neighborhood value depends on different experimental environments. Finally, the average of the points in each cluster is calculated to obtain the first coordinates of the center points in the camera image coordinate system. It should be noted that since the first projected image is a light spot and the finally recognized is the center point, the set of center points is sparse in the camera image coordinate system.

[0135] The embodiments of the present application can accurately identify the center points in the first camera image and obtain their corresponding first coordinates through precise edge detection and clustering analysis, ensuring the accuracy and flexibility of center point recognition.

[0136] In some embodiments, the above step S1112 includes the following steps:

[0137] Step S311: Determine the second coordinates of the calibration points based on the spatial relationship of the calibration points in the first projected image.

[0138] The first projected image is a set of multiple dot matrix images. To determine the second coordinates of each calibration point in each dot matrix image, the second coordinates of the first point in a dot matrix image in the projected image coordinate system can be obtained, and the second coordinates of all calibration points in the first projected image in the projected image coordinate system are determined based on the spatial relationship between each dot matrix element (i.e., calibration point) in the dot matrix image. Specifically, the above spatial relationship is that the column interval between each point in the dot matrix image is c , and the row interval is r .

[0139] As a specific example, assume that the first point in the m th dot matrix image is , then the point in the i th column j th row of this dot matrix image can be expressed as .

[0140] Based on the above spatial relationship, the second coordinates of all calibration points can be determined for subsequent determination of the mapping relationship between the calibration points and the center points.

[0141] Referring to Figure 4 shown: In some embodiments, the above step S1113 includes the following steps:

[0142] Step S321: Generate a first mapping table between the center points and the calibration points based on the first mapping relationship between the first coordinates and the second coordinates;

[0143] Step S322: When the number of center points is less than or equal to the number of calibration points, determine the repeated mapping points in the projection image coordinate system based on the first mapping table and the second coordinates;

[0144] Step S323: Determine a second mapping table between the center points to be complemented and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points;

[0145] Step S324: Generate a target mapping table based on the first mapping table and the second mapping table;

[0146] Wherein, the repeated mapping points include at least two calibration points having a first mapping relationship with the same center point.

[0147] Exemplarily, the second coordinate of the i column j row point in the above dot matrix image can be expressed as , based on the first camera image and the set of center points recognized, it can be determined that the first coordinate of the center point corresponding to this point in the first camera image can be expressed as , and the first mapping relationship between the first coordinate and the second coordinate can be expressed as . In step S321, based on this first mapping relationship, a first mapping table between the center points and the calibration points can be generated.

[0148] It should be understood that since the projection area of the A-pillar 304 is an irregular curved surface with discontinuous and inconsistent curvatures, and at the same time, the projector device may be affected by objective factors such as temperature during the projection process, the first mapping relationships between the center points and the calibration points are different from each other, and all the first mapping tables between the center points and the calibration points can be represented as a set .

[0149] Affected by the error of the image captured by the camera and the error of the center point recognition, the number of elements in the set of center points is less than or equal to the number of elements in the first projection image dot matrix set , that is, multiple calibration points of a part of the first projection image dot matrix set correspond to the same center point in the first camera image, and they have a first mapping table relationship. In step S322, based on the first mapping table and the set of second coordinates corresponding to the dot matrix set of the first projection image , the set of repeated calibration points from the first projection image dot matrix set in the projection image coordinate system can be determined, where RIt is a set of all repeated calibration points (i.e., repeated mapping points).

[0150] Due to the existence of repeated mapping points, there is no one-to-one correspondence between the center points and the calibration points. Therefore, it is necessary to complete these repeated mapping points. In step S323, based on the repeated mapping points and the center points corresponding to the repeated mapping points, a second mapping table between the center points to be completed and the repeated mapping points can be determined , where F 2 is a set of the second mapping table.

[0151] In step S324, by merging the first mapping table and the second mapping table , the target mapping table is obtained, where F is a set of the target mapping table.

[0152] Through the above steps, the target mapping relationship between all center points and calibration points can be obtained.

[0153] In some embodiments, the above step S323 includes the following steps:

[0154] Step S331: Based on the first coordinate of the center point to be completed, use a convolution kernel to scan the first camera image to generate a third coordinate of the center point to be completed in the camera image coordinate system;

[0155] Step S332: Based on the second coordinate and the third coordinate, generate a second mapping table between the center point to be completed and the repeated mapping points.

[0156] In the process of completing the center points of the repeated mapping points, a position-based convolution interpolation method can be adopted. It should be understood that the position-based convolution interpolation creates a new mapping relationship based on the neighborhood of the center point corresponding to the repeated mapping point, so as to obtain the center point to be completed.

[0157] Specifically, with the center point corresponding to the repeated mapping point as the center, based on k × k the first coordinates of the center points within the neighborhood of the size, use a convolution kernel of size k × k to scan and calculate the new camera image coordinate system coordinates. The number of repeated points within this neighborhood is r , and the convolution kernel weights are:

[0158] ; where w n is the weight corresponding to each neighborhood point, ([[]] x n , yn ) are the coordinates of neighboring points. Therefore, the third coordinate of the center point to be completed can be obtained by the following formula:

[0159] .

[0160] It should be noted that while the convolutional kernel scans the first camera image, the set R of repeated mapping points needs to be updated to prevent the third coordinate of the center point to be completed from still being repeated after convolutional calculation.

[0161] After completion, in step S332, a second mapping table is generated between the center point to be completed and the repeated mapping points , and there is a one-to-one correspondence between the center point to be completed and the repeated mapping points.

[0162] In some embodiments, the above step S323 includes the following steps:

[0163] Step S341: Based on the first coordinate of the center point to be completed, shift the first coordinate of the center point to be completed in the first direction by a predetermined number of pixels to generate a fourth coordinate of the center point to be completed in the camera image coordinate system;

[0164] Step S342: Based on the second coordinate and the fourth coordinate, generate a second mapping table between the center point to be completed and the repeated mapping points.

[0165] In the process of completing the center points of the repeated mapping points, the pixel offset method can also be used. Specifically, the pixel offset method recursively shifts the center point corresponding to the repeated mapping point one pixel to the right until the first left side of the center point corresponding to the repeated mapping point no longer coincides, so as to form a one-to-one mapping relationship between the center point and the correction point.

[0166] It should be understood that the first direction can be to the right or to the left, and the predetermined number of pixels can be 1 pixel, or 2 or even more pixels, which can be adjusted according to actual usage requirements during use, and the present application does not make any limitations in this regard.

[0167] After completion, in step S342, a second mapping table is generated between the center point to be completed and the repeated mapping points , and there is a one-to-one correspondence between the center point to be completed and the repeated mapping points.

[0168] Using the above pixel offset method scheme for center point completion, the algorithm is simple and the calculation time is short, having a certain advantage in processing efficiency and being suitable for scenarios that require fast processing. While using the above convolutional interpolation method for center point completion, the algorithm is more complex compared to the pixel offset method, but the convolutional interpolation method takes into account the spatial relationship between the repeated mapping points and the center points corresponding to the repeated mapping points, providing a more accurate mapping relationship and being applicable to scenarios that require high precision.

[0169] Referring to Figure 5 as shown, in some embodiments, the above step S112 includes the following steps:

[0170] Step S1121: Obtain the regional coordinates of the projection area in the camera image coordinate system according to the target mapping table;

[0171] Step S1122: Process the regional coordinates through at least two distortion correction methods to obtain at least two corrected camera images;

[0172] Step S1123: Determine the target method according to at least two corrected camera images;

[0173] Wherein, the projection area is the area for displaying the first projection image on the irregular curved surface; the target method is one of at least two distortion correction methods.

[0174] After obtaining the complete target mapping table, a one-to-one mapping relationship is formed between the correction points and the center points. Next, the target method can be determined according to the target mapping table for image distortion correction.

[0175] In step S1121, by way of example, referring to Figures 3a to 3b as shown, based on the environmental configurations such as the cockpit environment of the vehicle 10, the driver's position, and the line of sight, a suitable projection area of the A-pillar 304 is selected, and the coordinates of this projection area in the camera image coordinate system are obtained according to the target mapping table. The shape of this projection area is a straight line fitting the left and right sides of the A-pillar 304, and it can be, including but not limited to, a rectangle, a parallelogram, etc. The specific shape is determined when adapting to the vehicle model. The selection of the projection area can use the four vertex coordinates (i.e., regional coordinates) of the regional geometric figure in the camera image coordinate system:

[0176] ;

[0177] Wherein ul , ll , ur and lr respectively represent the four vertex positions of the upper left, lower left, upper right, and lower right.

[0178] In step S1122, at least two of the distortion correction methods such as the perspective transformation method, the pixel matching method, the camera calibration method, the distortion model algorithm, and the multi-view geometric image algorithm are used to perform distortion removal processing on the above regional coordinates respectively, so as to obtain at least two corresponding corrected camera images respectively.

[0179] In step S1123, a preferred target method is determined based on at least two obtained corrected camera images. Exemplarily, the corrected camera images can be compared for the undistortion effect to select the distortion correction method corresponding to the corrected camera image with the best undistortion effect as the preferred target method.

[0180] In the embodiments of the present application, after distortion correction is performed by multiple distortion correction methods to determine the target method, the optimal target method can be selected for distortion correction in the case of the projection area of an irregular curved surface.

[0181] In some embodiments, the above step S1122 includes the following steps:

[0182] Step S401: Based on the target mapping table and the projection area, at least two corrected projection images are determined using at least two distortion correction methods, and at least two corrected camera images are obtained by using a camera to photograph at least two corrected projection images.

[0183] As a specific example, in the embodiments of the present application, the perspective transformation method and the pixel matching method are selected as two distortion correction methods to perform distortion correction on the first projection pattern to obtain two corrected projection images.

[0184] For ease of understanding, in the following description process, the projection image corrected by the perspective transformation method is the second projection image, and the second projection image photographed by the camera is the second camera image; the projection image corrected by the pixel matching method is the third projection image, and the third projection image photographed by the camera is the third camera image. The second projection image and the third projection image are both corrected projection images; the second camera image and the third camera image are both corrected camera images.

[0185] In step S1121, the regional coordinates of the projection area in the camera image coordinate system are obtained, and the coordinates of the points in the projection area are determined, that is, which points are included in the projection area. Specifically, based on four vertices ul , ll , ur and lr the regional boundary line of the projection area is obtained, and then all the points falling within the projection area can be obtained. The set of all points falling within the projection area is , where, ( X 1, Y 1) are the coordinates of the points in the projection area in the camera image coordinates. The points in this set can be used for the pixel matching method. Specifically, the pixel matching method directly uses the pixel set in the corresponding projection area geometric figure Projection is performed, and other pixels outside this set are set to black, that is, not displayed. The third projection image is projected using this pixel matching method. The third camera image obtained by camera shooting is undistorted in the camera image coordinate system and can be used as an alignment object.

[0186] Further, assume that there is a line-line correspondence between the first projection image and the projection geometric figure area, that is, the four sides of the first projection image maintain row-column correspondence with the edges of the projection area. It should be noted that different from the conventional perspective transformation which only uses four vertices for calibration, the curved surface non-linear characteristics of the A-pillar 304 area require using more matching point pairs for fitting to ensure the accuracy of the perspective transformation. Since the center points in the first camera image are sparse in the camera image coordinate system space, in order to obtain the above line-line correspondence, a suitable center point fitting method needs to be used to determine the geometric figure of the projection area. In the embodiment of the present application, the Euclidean distance fitting method is used, and equidistant sampling is performed N p a number of points in each row of the geometric figure, and equidistant sampling is performed M p a number of rows to obtain a set of sampled points . For each sampled point ( X p , Y p ), the center point with the smallest Euclidean distance from it is selected to ensure the accuracy of fitting. The calculation formula of the Euclidean distance is as follows:

[0187] ;

[0188] where, ( X i , Y i ) are the center points in the first camera image.

[0189] In this way, a set of points in the fitted camera image coordinate system can be obtained, and further a set of geometric figure points of the projection area required for perspective transformation can be obtained. In this embodiment, the number of sampled points N p in each row is 5 to ensure sufficient accuracy and coverage. The present application does not limit this.

[0190] Perspective Transformation is a geometric transformation that projects an image from one perspective to another, and can simulate the projection effect in three-dimensional space. Its core is to perform coordinate mapping on each pixel in the image through a 3×3 transformation matrix, so that the originally tilted and distorted image area is "straightened" or adjusted to a specific perspective.

[0191] The perspective transformation preserves the straight-line shape of the image but changes the parallelism of the straight lines in the image compared to the affine transformation. The transformation formula for homogeneous coordinates of the perspective transformation is as follows:

[0192] ;

[0193] where the left side of the equation is the transformed coordinates, and the right side is the transformation matrix T and the coordinates before transformation; a 11 , a 12 , a 21 , a 22 control rotation, scaling, and shear (the affine transformation part); a 13 , a 23 control translation; a 31 , a 32 control the perspective effect, which is the core of the perspective transformation.

[0194] The transformed coordinates are rewritten as:

[0195] ;

[0196] Then the perspective transformation formula becomes:

[0197] .

[0198] where in the transformation matrix a 33 term can be rewritten as 1 to obtain a new transformation matrix:

[0199] .

[0200] Therefore, the number of unknowns in this transformation matrix changes from 9 to 8. Since 1 set of matching point pairs provides 2 equations, in traditional projection correction, 4 sets of matching point pairs are required for the perspective transformation, that is, the 4 vertices of the correction area; however, to better eliminate the influence of the irregular curved surface of the A-pillar 304, the embodiments of the present application use more point pairs, that is, the above N p × M p points. To obtain the point pairs in the projection image coordinate system corresponding to the fitting point set similarly, equally spaced row and column sampling is performed in the first projection image N p × M p points, and finally the transformation matrix Tp Based on the transformation matrix T p and the first projection image, the second projection image can be calculated, that is, the target projection image pre-distorted by the perspective transformation method, and the projected image is the corrected image, and most of the distortion has been removed.

[0201] The second camera image is obtained by using the camera to photograph the second projection image. Due to the irregular curved surface characteristics of the A-pillar 304, the second camera image is a distorted image.

[0202] The advantage of the pixel matching method is that it can present the projection image without distortion and maximize the image quality of the visualization of the A-pillar 304. However, the projectable area is limited and is a pixel set in a fixed area. .

[0203] The perspective transformation method can project images in any area or range, maximizing the retention of blind area information. However, its visualization effect is related to the camera position and the design of the A-pillar 304, and the projected image still has some distortion.

[0204] Therefore, exemplarily, when matching the models with functions, it is necessary to determine which method to use according to the effects of the two. Through the above perspective transformation method and pixel matching method, two corrected projection images can be obtained: the second projection image and the third projection image, and two corrected camera images: the second camera image and the third camera image. These corrected camera images will be used later to determine the target method for distortion correction.

[0205] In some embodiments, the above step S1123 includes the following steps:

[0206] Step S411: Determine the degree of coincidence of at least two corrected camera images;

[0207] Step S412: Determine the target method according to the value of the degree of coincidence;

[0208] Among them, the degree of coincidence is determined according to the overlapping area between at least two corrected camera images.

[0209] RMSE (Root Mean Square Error) is used to measure the degree of difference between two images. It evaluates the degradation of image quality by calculating the square root of the average of the squared differences of the corresponding pixel points of the two images. The smaller the RMSE value, the more similar the two images are, which is widely used in image quality evaluation, such as for quantifying image reconstruction, image compression, or the accuracy of prediction models.

[0210] However, in the specific scenario involved in the embodiments of the present application, due to the differences in the projected images, the traditional RMSE method is not suitable for evaluating the effects of two corrected camera images. Therefore, exemplarily, a comparison method is adopted in the embodiments of the present application to evaluate the quality of the second camera image obtained by the perspective transformation method.

[0211] Specifically, the method includes the following steps:

[0212] Perform binarization processing on the second camera image and the third camera image respectively. Binarization is an image processing technique that converts the pixel values in an image into a simple form of 0 or 1 (or black and white) for subsequent comparison and analysis.

[0213] Compare the binarized second camera image and the third camera image, and calculate the coincidence degree (coincidence rate) of their effective pixels (i.e., non-zero pixels). This step evaluates their similarity by comparing the matching conditions of the effective pixels at the same positions in the two images.

[0214] Determine whether to use the perspective transformation method as the final correction method according to the comparison result of the value of the coincidence rate and the preset threshold. In this embodiment, the preset threshold is set to 3% of the total number of camera pixels. If the value of the coincidence rate of the effective pixels is lower than the preset threshold, it indicates that the quality of the second camera image is acceptable, and the perspective transformation method can be used as the final correction method; otherwise, if the value of the coincidence rate is higher than the preset threshold, it indicates that the perspective transformation method may not be suitable, and at this time, the pixel matching method should be used for correction.

[0215] This method can more accurately reflect the quality of the image after perspective transformation by directly comparing the coincidence of the effective pixels, thereby providing a basis for selecting a suitable correction method.

[0216] The correction method provided by the embodiments of the present application focuses on the visualization scenario of the cockpit A-pillar 304 compared with the existing projector distortion correction methods. It uses a camera to simulate the position and field of view of the human eye, saves relevant parameters after correction, and the camera can be removed, that is, it is used when developing functions and matching models, and no additional hardware devices are required during mass production. It has a lower cost compared with the existing methods such as built-in cameras, TOF (Time of Flight), and gyroscopes, and is more extreme in cost reduction. At the same time, the embodiments of the present application consider that traditional trapezoidal correction or projector correction methods mostly use methods such as affine transformation, perspective transformation, plane fitting, and pose estimation, and the effect deteriorates when dealing with a curved surface with an irregular curvature of the projection surface. Under different vehicle models and different cockpit designs, the degree of deterioration of the perspective transformation method is different. Therefore, an additional pixel matching method is used for comparison and verification and the optimal one is selected.

[0217] Through the above technical solutions, an appropriate distortion correction method can be selected to perform projector projection distortion correction, effectively solving the problem of linear or non-linear distortion of the projection screen when the projector projects onto an irregular curved surface related to the vehicle, thereby improving the aesthetic degree of the projection screen relative to the driver's field of view and enhancing the user experience.

[0218] According to the second aspect of the present application, an embodiment of the present application further provides a correction system 300, which is applied to a vehicle 10, as Figure 6 shown. The system includes:

[0219] An in-vehicle projector 303, configured to project a target projection image onto an irregular curved surface related to the vehicle 10;

[0220] A distortion correction module 305, configured to perform distortion correction on the target projection image by using the correction method described in any of the above embodiments.

[0221] As a specific example, in the embodiment of the present application, the vehicle model is selected as the fixed model Song PLUS. The camera is installed at the position of the driver's common seat, and the height of the camera is 1.2 m, which is the average height of the driver's sitting posture. The camera has a sufficient field of view and can capture all projection screens. Therefore, the camera capture view is Figure 1 consistent with the human eye view; at the same time, the camera has a high resolution, and the captured first camera image is clear enough, which is more convenient for the recognition of the center point. Exemplarily, the camera in this embodiment includes, but is not limited to, using a Hikvision industrial camera, and the lens includes, but is not limited to, HF0628M-6MPE; the projector is installed in the center console, and the projection lens part is hollowed out and exposed. Exemplarily, the projector can be a MEMS (Micro-Electro-Mechanical Systems) projector, which is inexpensive and has a low hardware cost; Exemplarily, the projector can also be a DLP (Digital Light Processing) projector. Under the current industry process, the clarity of this type of projector is slightly higher, but the price is more expensive, and most of this type of projector belongs to the supplier Texas Instruments and is not suitable for independent research and development; Exemplarily, the projector can also be an LCOS (Liquid Crystal on Silicon) projector, which is more expensive. The present application does not limit the selection of the projector, and can be reasonably selected according to application requirements.

[0222] It should be noted that the projection area of the projector needs to be greater than or equal to the area of the A-pillar 304, so as to ensure that the corrected image can cover the area of the A-pillar 304.

[0223] It should be understood that this correction system 300 has all the beneficial effects of the above correction method, and the present application will not elaborate.

[0224] According to the third aspect of the present application, embodiments of the present application further provide a non-transitory computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the above correction method are implemented. This non-transitory computer-readable storage medium has all the beneficial effects of the above correction method, and the present application will not elaborate herein.

[0225] According to the fourth aspect of the present application, embodiments of the present application further provide a computer program product, including a computer program, and when the computer program is executed by a processor, the above correction method is implemented and has all the beneficial effects of the above correction method, and the present application will not elaborate herein.

[0226] According to the fifth aspect of the present application, embodiments of the present application further provide an electronic device, including: a memory and a processor, a computer program is stored on the memory; the processor is configured to execute the computer program in the memory to implement the steps of the above correction method. This electronic device has all the beneficial effects of the above correction method, and the present application will not elaborate herein.

[0227] The computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. The present application does not make specific limitations in this regard. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, 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 above.

[0228] In some embodiments of the present application, the computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, apparatus, or device.

[0229] The above computer-readable storage medium can be included in the above electronic device: or can exist separately without being assembled into the electronic device. The above computer-readable storage medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device is caused to:

[0230] Perform distortion correction on the target projection image using the target method;

[0231] Wherein, the target projection image is an image projected on an irregular curved surface related to a vehicle.

[0232] Computer program code for performing the operations of some embodiments of the present application may be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network (including a Local Area Network (LAN) or a Wide Area Network (WAN)), or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0233] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing a specified logical function.

[0234] It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings.

[0235] For example, two consecutively represented blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0236] The units described in some embodiments of the present application may be implemented in software or in hardware. The described units may also be provided in a processor.

[0237] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Parts (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), and so on.

[0238] According to the sixth aspect of the present application, as Figure 7 shown, an embodiment of the present application further provides a vehicle 10, including the above correction system 300 or an electronic device or implementing the correction method described in any of the above embodiments, and this electronic device can be used to execute the above correction method. The vehicle 10 has all the beneficial effects of the above electronic device, and the present application will not elaborate herein.

[0239] The vehicle 10 can be a fuel vehicle, a plug-in hybrid vehicle, a new energy vehicle, etc., and the present application does not make specific limitations thereto.

[0240] In the description of the present application, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of the present application, "a plurality of" means two or more, unless otherwise specifically defined.

[0241] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0242] Among the embodiments, implementation manners, and related technical features of the present application, they can be combined and replaced with each other without conflict.

[0243] The above are only the preferred embodiments of the present application and do not impose any form of limitation on the present application. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application still fall within the scope of the technical solution of the present application.

Claims

1. A calibration method, characterized in that, Applied to a vehicle, the method includes: Generating a target mapping table according to the center points in the first camera image and the corrected points in the first projection image; Obtaining the regional coordinates of the projection area in the camera image coordinate system according to the target mapping table; Processing the regional coordinates by at least two distortion correction methods to obtain at least two corrected camera images; Determining a target method according to at least two of the corrected camera images; Performing distortion correction on the target projection image using the target method; Wherein, the target projection image is an image projected on an irregular curved surface related to the vehicle; The first projection image is an image projected on the irregular curved surface; The first camera image is an image obtained by the camera photographing the first projection image; The projection area is the area on the irregular curved surface where the first projection image is displayed; the target method is one of at least two of the distortion correction methods; The center points are the points in the first camera image, and the corrected points are the points in the first projection image.

2. The method according to claim 1, wherein The irregular curved surface includes the surface of the vehicle's A-pillar.

3. The method according to claim 1 or 2, characterized in that, The target method includes one of a perspective transformation method, a pixel matching method, a camera calibration method, a distortion model algorithm, and a multi-view geometry image algorithm.

4. The method according to claim 1, wherein The camera is a human eye measurement camera installed on the vehicle.

5. The method according to claim 1, wherein The generating a target mapping table according to the center points in the first camera image and the corrected points in the first projection image includes: Obtaining the first coordinates of the center points in the first camera image in the camera image coordinate system; Obtaining the second coordinates of the corrected points in the first projection image in the projection image coordinate system; Generating the target mapping table between the center points and the corrected points based on the first coordinates and the second coordinates.

6. The method according to claim 5, characterized in that, Before obtaining the first coordinates of the center points in the first camera image in the camera image coordinate system, the method further includes: Performing image enhancement on the first camera image to identify the center points.

7. The method according to claim 5, characterized in that, The obtaining the first coordinates of the center points in the first camera image in the camera image coordinate system includes: Obtaining a set of detection points based on the first camera image; Using a clustering algorithm to determine the first coordinates; Wherein, the set of detection points is determined by identifying the dot matrix elements in the first camera image; The first coordinates are determined by performing clustering analysis on the set of detection points.

8. The method according to claim 5, characterized in that The obtaining the second coordinates of the corrected points in the first projection image in the projection image coordinate system includes: Determining the second coordinates of the corrected points based on the spatial relationship of the corrected points in the first projection image.

9. The method according to any one of claims 5 to 8, characterized in that The generating the target mapping table between the center points and the corrected points based on the first coordinates and the second coordinates includes: Generating a first mapping table between the center points and the corrected points based on the first mapping relationship between the first coordinates and the second coordinates; In the case where the number of the center points is less than or equal to the number of the corrected points, determining the repeated mapping points in the projection image coordinate system based on the first mapping table and the second coordinates; Determine a second mapping table between the center point to be completed and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points; Generate the target mapping table based on the first mapping table and the second mapping table; Wherein, the repeated mapping points include at least two calibration points having a first mapping relationship with the same center point.

10. The method according to claim 9, wherein The determining the second mapping table between the center point to be completed and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points includes: Based on the first coordinate of the center point to be completed, use a convolution kernel to scan the first camera image to generate a third coordinate of the center point to be completed in the camera image coordinate system; Generate the second mapping table between the center point to be completed and the repeated mapping points based on the second coordinate and the third coordinate.

11. The method according to claim 9, wherein The determining the second mapping table between the center point to be completed and the repeated mapping points based on the repeated mapping points and the center points corresponding to the repeated mapping points includes: Based on the first coordinate of the center point to be completed, offset the first coordinate of the center point to be completed in a first direction by a predetermined number of pixels to generate a fourth coordinate of the center point to be completed in the camera image coordinate system; Generate the second mapping table between the center point to be completed and the repeated mapping points based on the second coordinate and the fourth coordinate.

12. The method according to claim 1, characterized in that, The processing the regional coordinates by at least two distortion correction methods to obtain at least two corrected camera images includes: Based on the target mapping table and the projection area, use at least two of the distortion correction methods to determine at least two corrected projection images, and use the camera to capture at least two of the corrected projection images to obtain at least two corrected camera images.

13. The method according to claim 1, wherein The determining the target method according to at least two of the corrected camera images includes: Determine the coincidence degree of at least two of the corrected camera images; Determine the target method according to the value of the coincidence degree; wherein, the coincidence degree is determined according to the overlapping area between at least two of the corrected camera images.

14. A calibration system, applied to a vehicle, characterized in that, The system includes: A vehicle-mounted projector for projecting a target projection image on an irregular curved surface related to the vehicle; A distortion correction module for performing distortion correction on the target projection image using the correction method according to any one of claims 1 to 13.

15. A non-transitory computer-readable storage medium storing a computer program thereon, characterized in that, The computer program, when executed by a processor, implements the correction method according to any one of claims 1 to 13.

16. A computer program product comprising a computer program or instructions, characterized in that, The computer program or instruction, when executed by a processor, implements the correction method according to any one of claims 1 to 13.

17. An electronic device, characterized in that, Includes: A memory having a computer program stored thereon; A processor for executing the computer program in the memory to implement the correction method according to any one of claims 1 to 13.

18. A vehicle, characterized in that, Includes the correction system according to claim 14, or includes the electronic device according to claim 17.

Images