Vehicle-mounted camera calibration method and device, vehicle and storage medium

Abstract

The invention relates to a vehicle-mounted camera calibration method and device, a vehicle and a storage medium, and the method comprises the steps: obtaining environment images collected by a plurality of vehicle-mounted cameras disposed on the vehicle; splicing the environment images acquired by the plurality of vehicle-mounted cameras to generate a current panoramic image of the vehicle; and at least one abnormal feature is extracted from the current panoramic image of the vehicle, and the vehicle-mounted camera which is set abnormally is calibrated according to the at least one abnormal feature. Therefore, the technical problems that in the related technology, the requirement for the detection environment state is harsh, the abnormal installation problem of the vehicle-mounted camera cannot be visually displayed, the technical requirement for detection engineers is high, the detection cost is high, and application and popularization are difficult are solved.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent driving, and particularly relates to a calibration method, device, vehicle and storage medium for an in-vehicle camera. Background Art

[0002] There are consistency problems in the production process of in-vehicle cameras, and the in-vehicle cameras are installed on the vehicle at a certain angle and position, and there are also installation deviations.

[0003] In the functional development stage of intelligent driving technology, there is one or more in-vehicle cameras with incorrect positions, resulting in incorrect image output. Since the calibration function was not fully tuned in the early stage of development, it is difficult to accurately judge whether the installation position of the camera is correct offline.

[0004] In the related art, in the vehicle production line, a calibration board can be used to judge the installation position of the in-vehicle camera. For example, the specific corner positions of the checkerboard of the production line calibration board are restored by using the image captured by the in-vehicle camera. The core principle is to solve the geometric model parameters of camera imaging to ensure the accuracy of image information, and then calculate the values of internal and external parameters. Based on the external parameter calibration of the camera, the pose of the camera relative to the vehicle coordinate system is calculated, including translation and rotation information, and then specific points or features in the known world coordinate system are observed and mapped into the images collected by the in-vehicle camera to determine the pose of the camera.

[0005] However, in the related art, the environmental conditions are relatively harsh. It is necessary to ensure that the production line calibration function is fully developed and available; it is necessary to ensure that there is a four-wheel alignment system environment in the production line, and calibration tools such as checkerboards are configured, and the placement position of the checkerboard meets the image acquisition requirements of the vehicle camera, and the installation height and vehicle distance of the checkerboard meet the requirements. At the same time, it is difficult to intuitively judge whether there is a problem with the installation position of the in-vehicle camera in the related art. It is necessary for algorithm development engineers to accurately calculate the camera pose and then compare it with the designed position of the in-vehicle camera to judge whether the camera installation position is reasonable. The requirements for the working ability of inspection engineers are relatively high, and the labor cost is relatively high, which needs to be improved urgently. Summary of the Invention

[0006] The present application provides a calibration method, device, vehicle and storage medium for an in-vehicle camera to solve the technical problems in the related art that the requirements for the detection environment state are relatively harsh, the installation abnormality problem of the in-vehicle camera cannot be visually displayed, the technical requirements for inspection engineers are relatively high, the detection cost is relatively high, and it is difficult to promote and apply.

[0007] The first aspect of the embodiments of the present application provides a calibration method for an in-vehicle camera, including the following steps: obtaining environmental images collected by a plurality of in-vehicle cameras already installed on the vehicle; stitching the environmental images collected by the plurality of in-vehicle cameras to generate the current panoramic image of the vehicle; extracting at least one abnormal feature from the current panoramic image of the vehicle, and calibrating the in-vehicle camera with abnormal settings according to the at least one abnormal feature.

[0008] Optionally, in an embodiment of the present application, the extracting at least one abnormal feature from the current panoramic image of the vehicle includes: extracting at least one stitching area from the current panoramic image; determining a plurality of stitching features according to the plurality of stitched images corresponding to each stitching area, and judging whether the preset stitching condition is satisfied based on the plurality of stitching features; if the preset stitching condition is satisfied, generating an abnormal feature based on the plurality of stitching features.

[0009] Optionally, in an embodiment of the present application, the stitching the environmental images collected by the plurality of in-vehicle cameras to generate the current panoramic image of the vehicle includes: respectively performing image preprocessing on the environmental images collected by each in-vehicle camera to obtain a plurality of basic images that meet the preset quality conditions; extracting corresponding image features from each basic image based on the preset feature requirements to perform image registration using the image features and obtain corresponding registration results; performing image fusion based on the registration results to obtain the current panoramic image.

[0010] Optionally, in an embodiment of the present application, the extracting corresponding image features from each basic image based on the preset feature requirements includes: determining the type of feature to be extracted based on the preset feature requirements; matching the type of feature to be extracted with a corresponding feature detector, and setting a corresponding feature descriptor based on the feature detector; extracting the image features from each basic image using the feature detector and the feature descriptor.

[0011] Optionally, in an embodiment of the present application, it further includes: matching adjustment values of the in-vehicle camera with abnormal settings according to the at least one abnormal feature; adjusting the in-vehicle camera with abnormal settings according to the adjustment values of the in-vehicle camera with abnormal settings until the at least one abnormal feature is eliminated.

[0012] The second aspect of the embodiments of the present application provides a calibration device for an in-vehicle camera, including: an acquisition module, configured to acquire environmental images collected by a plurality of in-vehicle cameras already installed on the vehicle; a stitching module, configured to stitch the environmental images collected by the plurality of in-vehicle cameras to generate the current panoramic image of the vehicle; a calibration module, configured to extract at least one abnormal feature from the current panoramic image of the vehicle, and calibrate the in-vehicle camera with abnormal settings according to the at least one abnormal feature.

[0013] Optionally, in an embodiment of the present application, the calibration module includes: an extraction unit configured to extract at least one connection area from the current panoramic image; a determination unit configured to determine a plurality of stitching features according to a plurality of stitching images corresponding to each connection area, and determine whether a preset connection condition is satisfied based on the plurality of stitching features; a generation unit configured to generate an abnormal feature based on the plurality of stitching features when the preset connection condition is satisfied.

[0014] Optionally, in an embodiment of the present application, the stitching module includes: a preprocessing unit configured to perform image preprocessing on the environmental images collected by each vehicle-mounted camera respectively to obtain a plurality of basic images that meet the preset quality conditions; a registration unit configured to extract corresponding image features from each basic image based on a preset feature requirement, and use the image features for image registration to obtain corresponding registration results; a fusion unit configured to perform image fusion based on the registration results to obtain the current panoramic image.

[0015] Optionally, in an embodiment of the present application, the registration unit includes: a determination subunit configured to determine a feature type to be extracted based on the preset feature requirement; a matching subunit configured to match a corresponding feature detector with the feature type to be extracted, and set a corresponding feature descriptor based on the feature detector; an extraction subunit configured to extract the image features from each basic image by using the feature detector and the feature descriptor.

[0016] Optionally, in an embodiment of the present application, it further includes: a matching module configured to match the adjustment value of the vehicle-mounted camera with an abnormality set according to the at least one abnormal feature; an adjustment module configured to adjust the vehicle-mounted camera with an abnormality set according to the adjustment value of the vehicle-mounted camera with an abnormality set until the at least one abnormal feature is eliminated.

[0017] An embodiment of the third aspect of the present application provides a vehicle, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the calibration method of the vehicle-mounted camera as described in the above embodiment.

[0018] An embodiment of the fourth aspect of the present application provides a computer-readable storage medium, where the computer-readable storage medium stores computer instructions, and the computer instructions are used to cause the computer to execute the calibration method of the vehicle-mounted camera as described in the above embodiment.

[0019] An embodiment of the fifth aspect of the present application provides a computer program product, including a computer program, where when the computer program is executed, it is used to implement the calibration method of the vehicle-mounted camera as above.

[0020] Embodiments of the present application can obtain environmental images collected by multiple on-vehicle cameras installed on a vehicle, splice the environmental images collected by the multiple on-vehicle cameras to obtain a current panoramic image of the vehicle, and thus identify abnormal on-vehicle cameras according to abnormal features in the current panoramic image, so as to more intuitively display the installation abnormality problems of the on-vehicle cameras, reduce the technical requirements for detection engineers, and at the same time, have relatively low requirements for the detection environmental state, thereby reducing the overall detection cost and facilitating popularization and application. Thereby, the technical problems in the related art that the requirements for the detection environmental state are relatively harsh, the installation abnormality problems of the on-vehicle cameras cannot be intuitively displayed, the technical requirements for detection engineers are relatively high, the detection cost is relatively high, and it is difficult to promote and apply are solved.

[0021] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. Brief Description of the Drawings

[0022] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0023] Figure 1 is a flowchart of a calibration method for an on-vehicle camera according to an embodiment of the present application;

[0024] Figure 2 is a flowchart of a calibration method for an on-vehicle camera according to an embodiment of the present application;

[0025] Figure 3 is a schematic diagram of a panoramic image with a correctly installed on-vehicle camera according to an embodiment of the present application;

[0026] Figure 4 is a schematic diagram of a panoramic image with an incorrectly installed on-vehicle camera according to an embodiment of the present application;

[0027] Figure 5 is a schematic structural diagram of a calibration device for an on-vehicle camera according to an embodiment of the present application;

[0028] Figure 6 is a schematic structural diagram of a vehicle according to an embodiment of the present application. Detailed Embodiments

[0029] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as limiting the present application.

[0030] The calibration method, device, vehicle, and storage medium of an in-vehicle camera according to an embodiment of the present application will be described below with reference to the accompanying drawings. In the related art mentioned in the above background art, the requirements for the detection environment state are relatively harsh, and the installation abnormality problem of the in-vehicle camera cannot be visually displayed. The technical requirements for detection engineers are relatively high, the detection cost is relatively high, and it is difficult to promote and apply. The present application provides a calibration method for an in-vehicle camera. In this method, environmental images collected by a plurality of in-vehicle cameras already installed on the vehicle can be obtained, and the environmental images collected by the plurality of in-vehicle cameras can be stitched together to obtain the current panoramic image of the vehicle. Thus, abnormal in-vehicle cameras can be identified based on the abnormal features in the current panoramic image, so as to more intuitively display the installation abnormality problem of the in-vehicle camera, reduce the technical requirements for detection engineers, and at the same time, the requirements for the detection environment state are relatively low, thereby reducing the overall detection cost and facilitating promotion and application. Thereby, the technical problems in the related art that the requirements for the detection environment state are relatively harsh, the installation abnormality problem of the in-vehicle camera cannot be visually displayed, the technical requirements for detection engineers are relatively high, the detection cost is relatively high, and it is difficult to promote and apply are solved.

[0031] Specifically, Figure 1 is a schematic flowchart of a calibration method for an in-vehicle camera provided by an embodiment of the present application.

[0032] As Figure 1 shown, the calibration method of the in-vehicle camera includes the following steps:

[0033] In step S101, environmental images collected by a plurality of in-vehicle cameras already installed on the vehicle are obtained.

[0034] In order to solve the problems in the related art that when detecting the installation position of the in-vehicle camera of a vehicle on the production line, the requirements for the detection environment state are harsh, the skill knowledge requirements for the detection personnel are relatively high, and the intuitiveness of the detection result is relatively poor, etc., the embodiment of the present application can stitch the environmental images collected by a plurality of in-vehicle cameras and identify the in-vehicle camera with an abnormal installation position according to the stitching result.

[0035] In order to solve the above problems, the embodiment of the present application can obtain the environmental images collected by a plurality of in-vehicle cameras already installed on the vehicle to lay a foundation for subsequent image stitching.

[0036] When the on-vehicle camera in the embodiment of the present application collects environmental images, the super-exposure technology can also be used. This technology can separate the photosensitive and storage functions of pixels, and store the overflowing charges in a capacitor after the photosensitive diode is saturated, thereby significantly increasing the dynamic range of the pixels. This enables the image sensor to capture clear and accurate image details under various lighting conditions, including tunnel environments with strong contrast between light and dark or during night driving, enhancing the efficiency and accuracy of image fusion.

[0037] In step S102, the environmental images collected by multiple on-vehicle cameras are stitched to generate the current panoramic image of the vehicle.

[0038] Furthermore, in the embodiment of the present application, the environmental images collected by multiple on-vehicle cameras can be stitched to obtain the current panoramic image stitched from the environmental images, so as to determine the installation position of the on-vehicle camera based on the quality of the current panoramic image.

[0039] Among them, the current panoramic image can be displayed through the central control screen of the vehicle or through an external screen connected to the vehicle, so that relevant detection personnel can evaluate the quality of the current panoramic image to determine whether there is a deviation in the installation position of the on-vehicle camera.

[0040] The embodiment of the present application can also perform quality evaluation by identifying abnormal features in the current panoramic image, and then determine whether there is a deviation in the installation position of the on-vehicle camera. Specifically, it will be elaborated below.

[0041] Optionally, in an embodiment of the present application, stitching the environmental images collected by multiple on-vehicle cameras to generate the current panoramic image of the vehicle includes: respectively performing image preprocessing on the environmental images collected by each on-vehicle camera to obtain multiple basic images that meet the preset quality conditions; extracting corresponding image features from each basic image based on the preset feature requirements to perform image registration using the image features to obtain the corresponding registration results; and performing image fusion based on the registration results to obtain the current panoramic image.

[0042] As a possible implementation manner, during the image stitching process, the embodiment of the present application may include three steps: image preprocessing, image registration, and image fusion. To improve the quality of the stitched current panoramic image, the embodiment of the present application may also add an image post-processing step on the basis of the above steps to facilitate more accurate identification of abnormal features in the subsequent process.

[0043] Among them, in the image preprocessing stage, the embodiment of the present application can preprocess the collected environmental images to obtain basic images with qualified quality through distortion correction. Among them, the preset quality conditions can be set by those skilled in the art in combination with the current detection target, and no specific limitation is made here.

[0044] During the distortion correction process, the embodiments of the present application need to obtain multiple pieces of data, including:

[0045] Camera Intrinsics: including Focal Length: f x and f y ; Optical Center: c x and c y .

[0046] Distortion Coefficients: including Radial Distortion Coefficients: k1, k2 and / or k3, where k3 is a high-order radial distortion and can be selected according to the actual image; Tangential Distortion Coefficients: p1 and p2.

[0047] Image Size: the height and width of the image.

[0048] Generally, the image after distortion correction can be used for subsequent image registration. However, to reduce the influence of uneven illumination in the image, the embodiments of the present application can also perform illumination equalization on the environmental image to make the illumination conditions of each region of the image relatively consistent. The embodiments of the present application can use methods such as histogram equalization and adaptive histogram equalization for optimization.

[0049] Among them, histogram equalization includes global equalization (applying histogram equalization to the entire image) and local equalization (dividing the image into small blocks, applying histogram equalization to each small block separately, and then using interpolation methods to splice the small blocks back into a complete image).

[0050] Adaptive histogram equalization can introduce a limit parameter (the threshold of contrast limitation, used to clip the highest peak of the histogram and reduce noise amplification) and the number of blocks of the image (the size of the small blocks into which the image is divided, such as the grid size), and implement histogram equalization.

[0051] Gamma correction can be achieved through the Gamma value: a non-linear transformation parameter that controls the image brightness. A value less than 1 will brighten the image, and a value greater than 1 will darken the image.

[0052] However, in practical applications, it is necessary to select appropriate equalization methods and parameters according to the specific situation of the image. For example, histogram equalization cannot be selected for images with drastic illumination changes, which may lead to amplified image noise; adaptive histogram equalization can better handle local illumination, but it will increase the computational complexity.

[0053] In the image registration stage, the embodiments of the present application include two steps: feature extraction, feature matching, and transformation matrix calculation. Among them, feature extraction will be elaborated below.

[0054] In the feature matching stage, for example, the strategies of feature matching can include brute-force matching (Brute-Force Matcher), FLANN (Fast Library for Approximate Nearest Neighbors, an open-source library for approximate neighbors), etc. Specifically, it can be selected according to actual needs. Among them, when selecting the FLANN strategy, the internal parameters of FLANN can be controlled through index_params and search_params.

[0055] After determining the feature matching strategy, corresponding feature matching can be performed in combination with the matching threshold and screening conditions. Among them, the matching threshold is used to control the maximum distance or similarity between matching feature point pairs, and the screening conditions are used to remove incorrect matches, such as screening matching pairs through a ratio test.

[0056] In the process of calculating the transformation matrix, the embodiments of the present application may include the following steps:

[0057] Step S1: Determine the transformation type: Select an appropriate transformation type, such as affine transformation, perspective transformation (projection transformation), etc. Affine transformation preserves straight lines and parallelism, while perspective transformation can handle more complex deformations.

[0058] Step S2: Determine the number of matching points: A sufficient number of matching points are required to calculate the transformation matrix. The number and quality of matching points directly affect the accuracy of the transformation matrix.

[0059] Step S3: Robust algorithm: Use a robust algorithm to estimate the transformation matrix, such as RANSAC (Random Sample Consensus algorithm) or LMEDS (Least Median of Squares error), etc. These algorithms can handle situations with noise and incorrect matches.

[0060] Step S4: Error threshold: In algorithms such as RANSAC, set the reprojection error threshold to determine which points are considered inliers (conforming to the model) or outliers (not conforming to the model).

[0061] Step S5: Number of iterations: In algorithms such as RANSAC, a maximum number of iterations is set to control the balance between the running time and accuracy of the algorithm.

[0062] In the image fusion stage, embodiments of the present application can first determine the overlapping area. For example, ensure that the input images have the same size or at least partially overlapping sizes; if the images have been aligned through a certain transformation (such as affine transformation, perspective transformation), the transformation matrix can help determine the overlapping area; determine the threshold for which pixels belong to the overlapping area, usually based on color similarity or gradient change; determine the possible overlapping area by calculating the bounding box of the image.

[0063] Furthermore, embodiments of the present application can select a fusion strategy. For example, select a fusion strategy by image type: grayscale image or color image; infrared and visible light images, multi-spectral images, etc. Determine the fusion strategy by fusion objective: improve image clarity; enhance image contrast; retain important features (such as edges, textures), etc. The fusion methods can include: pixel-level fusion: directly operate on the pixel values of the image; feature-level fusion: first extract the features of the image and then perform fusion; decision-level fusion: perform fusion based on the classification or recognition results of the image; weight assignment: assign weights according to the quality, importance, or other criteria of the image; it can be fixed or adaptive. The fusion rules can include simple rules such as maximum value, minimum value, average value, weighted average value, etc.; more complex fusion rules, such as those based on multi-scale decomposition, sparse representation, deep learning, etc.

[0064] It should be noted that in the image fusion stage, it is essential to convert the color image to a grayscale image. Because directly fusing each channel of the color image may cause color distortion, in this case, more complex color space conversion and fusion strategies may be required. In the fusion strategy, the "sparse representation" strategy is preferred. Its principle is to decompose the image into two parts: sparse representation and dense representation. The sparse representation part contains high-frequency information such as textures and edges in the image, while the dense representation part contains low-frequency information in the image. By appropriately fusing the two parts, a clearer and richer image can be obtained. The advantages are: it can better retain the details and features of the image and has good extraction ability for high-frequency information in the image. The disadvantage of the "weighted average" fusion strategy is that the fusion effect is greatly affected by the weights, and improper weight selection will lead to a decline in the quality of the fused image; and it cannot completely retain all information, especially in areas such as image edges and details. The disadvantage of the "multi-scale decomposition" fusion strategy is that it may cause redundant decomposition of the image and may lose some high-frequency information during the decomposition process, resulting in more losses in the final fusion result.

[0065] In the image post-processing stage, embodiments of the present application can perform color correction and image enhancement.

[0066] Among them, color correction includes adjusting the overall brightness of the image, enhancing or weakening the difference between the bright and dark areas in the image, increasing or decreasing the vividness of the colors in the image, adjusting the white balance of the image to match different light source color temperatures, and changing the overall color tendency of the image, such as changing from a red tone to a blue tone, etc.

[0067] Image enhancement can be selected accordingly according to actual needs. For example, it can include:

[0068] Histogram equalization: Its control conditions can be without specific parameters, but it can be selected to operate on a certain channel or the entire image of the image. Its effect can be to enhance the contrast of the image and make the details in the image clearer.

[0069] Adaptive histogram equalization: Its parameters can be: clipLimit (contrast limit) and tileGridSize (the size of each small grid). Its effect can be to enhance local contrast while keeping the overall brightness of the image unchanged.

[0070] Sharpening: Its parameter can be the sharpening intensity (usually achieved through a convolution kernel). Its effect can be to enhance the image edges and make the image look clearer.

[0071] Denoising: Its parameters can be the type of denoising algorithm (such as Gaussian filtering, mean filtering, median filtering, bilateral filtering, etc.) and its parameters (such as the size of the filter, standard deviation, etc.). Its effect can be to reduce the noise in the image and improve the image quality.

[0072] Laplacian operator: Its parameter can be the size of the convolution kernel. Its effect can be used for edge detection and can also be used for image sharpening.

[0073] It should be noted that during the image post - processing stage, brightness adjustment, contrast adjustment, and saturation adjustment are essential. For brightness adjustment, the brightness is adjusted by increasing or decreasing the pixel values of the V (Value) channel, and note that the pixel values exceeding the range of 0 to 255 should be clipped. For contrast adjustment, a linear transformation is used to adjust the pixel values of the V channel, and the contrast control factor alpha varies from 0 to 2. When alpha = 1, the contrast remains unchanged; when alpha < 1, the contrast decreases; when alpha > 1, the contrast increases. For saturation adjustment, the saturation is adjusted by increasing or decreasing the pixel values of the S (Saturation) channel. Color temperature adjustment is preferred and usually involves more complex white - balance algorithms, and may require the use of a specific color mapping table or look - up table (LUT). Histogram equalization, adaptive histogram equalization, and sharpening are essential, and denoising is preferred. Histogram equalization has a good effect on images with low overall contrast, but may cause loss of details in some areas of the image. Adaptive histogram equalization can better preserve image details by restricting the contrast and performing histogram equalization within small regions. Using a sharpening convolution kernel can enhance image edges, but over - sharpening may cause noise or artifacts in the image. When selecting a denoising algorithm, it is necessary to select a suitable algorithm and its parameters according to the type and degree of noise in the image.

[0074] Optionally, in an embodiment of the present application, corresponding image features are extracted from each base image based on a preset feature requirement, including: determining the type of feature to be extracted based on the preset feature requirement; using the type of feature to be extracted to match the corresponding feature detector, and setting the corresponding feature descriptor based on the feature detector; using the feature detector and the feature descriptor to extract image features from each base image.

[0075] For example, in the embodiments of the present application, the types of features to be extracted can be selected first, such as point features (e.g., corner points, edge points), line features, region features, or global features (e.g., color histograms, texture features), etc. Further, in the embodiments of the present application, a feature detector can be selected, such as SIFT (Scale-Invariant Feature Transform), SURF (Speeded Up Robust Features), ORB (Oriented FAST and Rotated BRIEF), etc. Corresponding parameters need to be set, such as scale space parameters, edge thresholds, non-maximum suppression parameters, etc. On this basis, the embodiments of the present application can set corresponding feature descriptors to extract image features. Among them, the feature descriptor describes the local information of the feature points, such as SIFT descriptors, BRIEF descriptors, etc. Different parameters need to be set, such as the dimension of the descriptor, sampling mode, etc.

[0076] It should be noted that the preset feature requirements can be set accordingly according to the actual working conditions of the embodiments of the present application, and no specific limitations are made here.

[0077] In step S103, at least one abnormal feature is extracted from the current panoramic image of the vehicle, and the in-vehicle camera with an abnormality is calibrated according to the at least one abnormal feature.

[0078] In the actual execution process, the embodiments of the present application can extract abnormal features from the current panoramic image of the vehicle to determine whether there is a situation where the in-vehicle camera is installed in the wrong position.

[0079] Among them, the abnormal features can include various types, such as whether there are errors, distortions, or unnatural seams in the stitched image. Through the abnormal features, the embodiments of the present application can determine the specific in-vehicle camera with an installation error and determine the deviation from the correct installation position for position adjustment.

[0080] Optionally, in an embodiment of the present application, extracting at least one abnormal feature from the current panoramic image of the vehicle includes: extracting at least one stitching area from the current panoramic image; determining a plurality of stitching features according to the plurality of stitched images corresponding to each stitching area, and determining whether the plurality of stitching features meet a preset stitching condition based on the plurality of stitching features; if the preset stitching condition is met, generating an abnormal feature based on the plurality of stitching features.

[0081] In some embodiments, at least one connection area can be extracted from the current panoramic image, and then the connection area can be determined to determine whether the picture at the splicing point is complete and smoothly connected. For example, in the embodiments of the present application, objects in the current panoramic image can be used as reference objects, such as lanes, road edges, roadside trees, etc., to determine whether the shape of the reference object conforms to general rules, and use it as a splicing feature, and determine whether the splicing feature meets the connection conditions, such as whether the lane lines are not on the same straight line, whether the road edges are smooth, etc. If there are situations such as the lane lines being disconnected in the middle, the two ends being unable to align, or the road edges being not smooth, it can be determined that the preset connection conditions are met, and then it can be determined that the current panoramic image has abnormal features.

[0082] In the embodiments of the present application, simulated grid lines can also be added to the current panoramic image to use the grid lines for auxiliary judgment to determine whether the reference object in the current panoramic image is reasonable, that is, whether the splicing feature meets the preset connection conditions.

[0083] Based on the above method, the abnormality in the current panoramic image can be accurately identified, thereby reducing the professional technical requirements for the detection personnel.

[0084] In addition to the above solutions, it can be understood that when the vehicle performs image acquisition, there may be no obvious reference objects around. In this regard, in the embodiments of the present application, the vehicle itself can also be used as a reference object, that is, the vehicle structure is identified in the current panoramic image, and it is matched according to the vehicle model information to determine whether the vehicle structure is reasonable in the current panoramic image, and then abnormal features are extracted.

[0085] In the actual application process, during the pre-delivery inspection of the vehicle, image acquisition can be performed by driving the vehicle to the detection position (where multiple reference objects can be set around the detection position), and the detection personnel can determine the abnormal features by comparing the actual environment with the current panoramic image.

[0086] In addition to the above manual determination, automatic recognition determination can also be performed. For example, the consistency of feature point matching is detected in the overlapping area of on-vehicle cameras, and the unmatched on-vehicle cameras are found; for another example, the external parameters of the on-vehicle cameras are estimated during the splicing process and compared with the preset correct external parameters to identify the on-vehicle cameras with abnormal settings; for another example, continuous images and abnormal images are used as samples to train a related abnormal feature recognition model to automatically output abnormal features according to the current panoramic image, etc. In the actual execution process, corresponding settings can be made according to the actual type and configuration of the vehicle, which will not be specifically limited here.

[0087] Optionally, in an embodiment of the present application, it further includes: matching at least one abnormal feature to set an adjustment value for an abnormal vehicle-mounted camera; adjusting the abnormal vehicle-mounted camera according to the set adjustment value of the abnormal vehicle-mounted camera until at least one abnormal feature is eliminated.

[0088] In the embodiment of the present application, after extracting abnormal features, the adjustment value of the abnormal vehicle-mounted camera corresponding to the abnormal features can be matched, such as determining the adjustment value based on the angular deviation between the reference object and the grid line, so as to adjust the abnormal vehicle-mounted camera according to the adjustment data, and collect and splice images here until no abnormal features can be extracted from the obtained panoramic image.

[0089] In addition, in addition to being applied to vehicle production line detection, the embodiment of the present application can also perform vehicle self-check based on the calibration method of the embodiment of the present application after the vehicle is put into use.

[0090] It can be understood that after the vehicle is put into use, the position of the vehicle-mounted camera will be offset due to external forces, such as collisions and extreme weather.

[0091] Drivers do not have the relevant knowledge of professional technicians, so it is difficult to detect problems with vehicle-mounted cameras in a timely manner. When using image data collected by vehicle-mounted cameras with abnormal positions, the actual presented images will not match the driver's expectations, posing certain safety hazards.

[0092] The embodiment of the present application can realize self-check of the installation position of the vehicle-mounted camera through the above technical solutions, so as to timely remind the driver when an abnormality of the vehicle-mounted camera is detected.

[0093] In the actual execution process, in addition to the above detection methods, the embodiment of the present application can also collect images through multiple vehicle-mounted cameras before powering off after each parking of the vehicle, and splice and save the collected images. After the next power-on, call multiple vehicle-mounted cameras again to collect images for splicing, and determine whether the position of the vehicle-mounted camera has shifted during parking by comparing the spliced images after power-on and the spliced images when powering off, or continuously collect images during vehicle driving after the vehicle senses body vibration, and select images at the same moment for splicing, and then perform determination on the spliced panoramic image.

[0094] The vehicle can also access the weather forecast to detect whether the position of the vehicle-mounted camera has shifted by using the spliced panoramic image after extreme weather occurs.

[0095] After receiving the reminder of position deviation, the driver can make self-adjustment according to the adjustment instructions of the vehicle's vehicle-mounted camera, or contact relevant technical personnel for vehicle maintenance to ensure driving safety.

[0096] Combination Figures 2 to 4 As shown, the working principle of the calibration method for an in-vehicle camera according to an embodiment of the present application will be described in detail with an example.

[0097] In the actual execution process, taking the 7V (front wide-angle camera, front narrow-angle camera, left front camera, right front camera, left rear camera, right rear camera, rear camera) camera as an example, it is assumed that the vehicle to be detected is on a straight multi-lane road about 50m long, and there are obvious solid or dashed lane lines on the road, the road edge is complete without damage and perpendicular to the lane lines.

[0098] As Figure 2 shown, the embodiment of the present application may include the following steps:

[0099] Step S201: Collect images. The embodiment of the present application can control the vehicle to drive to the above-mentioned lane, determine that the 7V camera can normally output the video stream to the in-vehicle screen for display, and control the vehicle to drive along the road at a speed of 30 kph. Use the 7V camera to collect images, collect continuous 1-minute data packets, and then perform distortion preprocessing on a single image, read the image, and obtain the image size.

[0100] Step S202: Image stitching. The embodiment of the present application can use cv2.getOptimalNewCameraMatrix to obtain an optimal new camera matrix and a cropping region (ROI), so as to remove the black edges that appear at the image edges caused by distortion correction, correct the radial and tangential distortions in the image, and obtain a more real and accurate image. Then, read the image, use global histogram equalization to reduce the influence of uneven illumination in the image; create a CLAHE object, apply adaptive histogram equalization to better handle local illumination changes; construct a lookup table, apply the lookup table for gamma correction to adjust the overall brightness of the image.

[0101] Then, perform image registration to align the images obtained from different perspectives of the 7V camera at the same time to the same reference coordinate system. Next, read the image, initialize the SIFT feature detector, created by cv2.SIFT_create(); then, initialize the FLANN matcher through cv2.FlannBasedMatcher(), give the matching descriptors, apply the ratio test to filter the matching pairs, and filter high-quality matches by comparing the distance between each matching point and its second-best matching point. Draw the matching result, obtain the coordinates of the matching points, calculate the perspective transformation matrix through cv2.findHomography(), and use the RANSAC algorithm to handle potential incorrect matches. Transform the query image to the coordinate system of the training image. Apply perspective transformation for image registration and display the registration result.

[0102] Next, combine the information of 7 images at the same moment from different perspectives to generate a new image that contains the important features of all input images, namely the stitched image. First, the 7 input images are grayscale images. Then, create a non-zero pixel area as an image mask, and then find the intersection of the image masks, which is the overlapping area of the two images. Calculate the bounding box of the overlapping area and return its coordinate values. Next, perform the fusion of sparse representation. Through appropriate fusion of each part, a clearer and richer image can be obtained.

[0103] Finally, in the image post-processing stage, convert the fused image to the HSV color space, adjust the image brightness, contrast, and saturation in sequence, and then convert the image back to BGR. Then, perform operations such as histogram equalization, adaptive histogram equalization, and sharpening, and finally display the result.

[0104] Step S203: Determine whether there is a position deviation in the vehicle-mounted camera. The panoramic image obtained after stitching can be directly displayed on the vehicle side, such as being displayed under the 7V camera image diagnosis in the vehicle head unit - intelligent driving APK - debugging interface.

[0105] The effect of stitching of the 7V camera is displayed on the vehicle head unit screen, such as Figure 3 and Figure 4 as shown, where Figure 3 is the panoramic image when the installation position of the 7V camera is correct, and Figure 4 is the panoramic image when the installation of the left rear-view camera is incorrect.

[0106] The corresponding field of view of the front wide-angle camera is virtually 110° (within the red included angle area), and the corresponding field of view of the front narrow-angle camera is virtually 30° (within the orange included angle area). The areas shown are in the upper half of the picture, accounting for about 2 / 3 of the area; the corresponding fields of view of the left front camera, right front camera, left rear camera, and right rear camera are virtually 99° (within the yellow, green, cyan, and blue included angle areas respectively), covering most of the area, leaving only about 1 / 5 of the lower half area not fully covered. The corresponding field of view of the rear-view camera is virtually 50° (within the purple included angle area), complementing and exceeding the left lower half area by about 1 / 5. By comparing these line segments and angles of different colors, the picture can be divided into 14 areas of different sizes. With the help of a 1cm * 1cm red grid (actual size 10m * 10m), the specific positions and sizes of each area can be accurately located. If there is an incorrect installation of the camera position, the lane lines and road edges will have obvious deflections. At the same time, the sizes of the 14 areas will also change, either increasing or decreasing. Based on the presentation of the lane lines and surrounding characteristic trees / buildings after the fusion of the 7V camera, determine the installation position of the camera.

[0107] Taking the top view after the integration of a 7V camera to show the lane lines as an example. As Figure 3 shown, it can be found that the road edge and the lane are both perpendicular or parallel to the red grid in the figure, and the spliced images are complete and smoothly connected. It is relatively easy to judge that the camera is correctly installed. As Figure 4 shown, in the area circled by the red circle, there is a bend in the road edge, and the smoothness of the spliced images is poor. This situation appears in the lower left corner of the entire picture, and it is relatively easy to judge that it is caused by the deviation of the installation position of the left rear camera. In the area circled by the red rectangle, it can be found that there is an angular deviation between the lane line and the vertical red grid line, forming an angle of about 5°. This indicates that there is a deviation in the installation position of the left front camera. Since Figure 4 the right lane lines in it are all perpendicular or parallel to the red grid line, it shows that the positions of the forward wide-angle camera and the narrow-angle camera are both correctly installed. This picture is presented on the in-vehicle device and can be opened by entering from the panoramic APK view. Even for non-technical personnel, they can quickly judge whether the camera is installed incorrectly through the splicing effect.

[0108] In summary, the embodiment of the present application can break away from the original method of judging the camera position through a calibration board, quickly and directly complete the multi-camera image splicing, and use post-processing means such as image quality enhancement technology, such as color temperature optimization and denoising, all of which are to enhance the visual effect of the image and facilitate more intuitive judgment. The embodiment of the present application can also divide the spliced top view, that is, the panoramic image, into multiple different regions according to the actual positions of the vehicle-mounted cameras through the in-vehicle device, and focus on the display forms of reference objects such as lane lines and road edges in the splicing overlapping region. If there are obvious deflections or obvious angles with the vertical grid line, it indicates that there is a position deviation of the camera in a certain region. This reduces the need for manual intervention by professional technical personnel. Non-technical personnel can also find the detection trigger port through the in-vehicle intelligent driving APK, perform image diagnosis and judgment, and visually check the splicing effect with the naked eye to draw a conclusion on whether the camera is installed incorrectly. This greatly improves the accuracy and efficiency of judgment. And by introducing an automated processing algorithm, the image can be analyzed, processed, and optimized in real time, reducing the time of manual image acquisition and calculation model in the production line, and at the same time, improving the efficiency and accuracy of image processing.

[0109] According to the calibration method of the vehicle-mounted camera proposed by the embodiments of the present application, environmental images collected by a plurality of vehicle-mounted cameras already installed on the vehicle can be obtained, and the environmental images collected by the plurality of vehicle-mounted cameras can be stitched to obtain the current panoramic image of the vehicle, so as to identify abnormal vehicle-mounted cameras according to the abnormal features in the current panoramic image, more intuitively display the installation abnormality problems of the vehicle-mounted cameras, reduce the technical requirements for inspection engineers, and at the same time, have lower requirements for the detection environmental state, thereby reducing the overall detection cost and facilitating popularization and application. Thus, the technical problems in the related art that the requirements for the detection environmental state are relatively harsh, the installation abnormality problems of the vehicle-mounted cameras cannot be visually displayed, the technical requirements for inspection engineers are relatively high, the detection cost is relatively high, and it is difficult to popularize and apply are solved.

[0110] Next, a calibration device for a vehicle-mounted camera according to an embodiment of the present application will be described with reference to the accompanying drawings.

[0111] Figure 5 It is a block diagram of a calibration device for a vehicle-mounted camera according to an embodiment of the present application.

[0112] As Figure 5 shown, the calibration device 10 for the vehicle-mounted camera includes: an acquisition module 100, a stitching module 200, and a calibration module 300.

[0113] Specifically, the acquisition module 100 is configured to acquire environmental images collected by a plurality of vehicle-mounted cameras already installed on the vehicle.

[0114] The stitching module 200 is configured to stitch environmental images collected by a plurality of vehicle-mounted cameras to generate a current panoramic image of the vehicle.

[0115] The calibration module 300 is configured to extract at least one abnormal feature from the current panoramic image of the vehicle and calibrate the vehicle-mounted camera with an abnormal setting according to the at least one abnormal feature.

[0116] Optionally, in an embodiment of the present application, the calibration module 300 includes: an extraction unit, a judgment unit, and a generation unit.

[0117] Among them, the extraction unit is configured to extract at least one connection area from the current panoramic image.

[0118] The judgment unit is configured to determine a plurality of stitching features according to a plurality of stitched images corresponding to each connection area, and judge whether the preset connection condition is satisfied based on the plurality of stitching features.

[0119] The generation unit is configured to generate an abnormal feature based on the plurality of stitching features when the preset connection condition is satisfied.

[0120] Optionally, in an embodiment of the present application, the stitching module 200 includes: a preprocessing unit, a registration unit, and a fusion unit.

[0121] Among them, the preprocessing unit is configured to perform image preprocessing on the environmental images collected by each vehicle-mounted camera respectively, so as to obtain a plurality of basic images that meet the preset quality conditions.

[0122] The registration unit is configured to extract corresponding image features from each basic image based on the preset feature requirements, so as to perform image registration using the image features and obtain corresponding registration results.

[0123] The fusion unit is configured to perform image fusion based on the registration results, so as to obtain the current panoramic image.

[0124] Optionally, in an embodiment of the present application, the registration unit includes: a determination subunit, a matching subunit, and an extraction subunit.

[0125] Among them, the determination subunit is configured to determine the type of feature to be extracted based on the preset feature requirements.

[0126] The matching subunit is configured to match the corresponding feature detector using the type of feature to be extracted, and set the corresponding feature descriptor based on the feature detector.

[0127] The extraction subunit is configured to extract image features from each basic image using the feature detector and the feature descriptor.

[0128] Optionally, in an embodiment of the present application, the calibration device 10 of the vehicle-mounted camera further includes: a matching module and an adjustment module.

[0129] Among them, the matching module is configured to match and set the adjustment value of the abnormal vehicle-mounted camera according to at least one abnormal feature.

[0130] The adjustment module is configured to adjust the abnormal vehicle-mounted camera according to the adjustment value of the abnormal vehicle-mounted camera until at least one abnormal feature is eliminated.

[0131] It should be noted that the foregoing explanation of the embodiment of the calibration method of the vehicle-mounted camera is also applicable to the calibration device of the vehicle-mounted camera in this embodiment, and will not be repeated here.

[0132] The calibration device for in-vehicle cameras proposed according to the embodiments of the present application can acquire environmental images collected by a plurality of in-vehicle cameras already installed on a vehicle, splice the environmental images collected by the plurality of in-vehicle cameras to obtain the current panoramic image of the vehicle, and thus identify abnormal in-vehicle cameras according to the abnormal features in the current panoramic image, so as to more intuitively display the installation abnormality problems of in-vehicle cameras, reduce the technical requirements for detection engineers, and at the same time, have relatively low requirements for the detection environment state, thereby reducing the overall detection cost and facilitating popularization and application. Thus, it solves the technical problems in the related art that the requirements for the detection environment state are relatively harsh, the installation abnormality problems of in-vehicle cameras cannot be intuitively displayed, the technical requirements for detection engineers are relatively high, the detection cost is relatively high, and it is difficult to popularize and apply.

[0133] Figure 6 The structural schematic diagram of the vehicle provided by the embodiment of the present application. The vehicle may include:

[0134] A memory 601, a processor 602, and a computer program stored on the memory 601 and executable on the processor 602.

[0135] When the processor 602 executes the program, it implements the calibration method for in-vehicle cameras provided in the above embodiments.

[0136] Further, the vehicle further includes:

[0137] A communication interface 603 for communication between the memory 601 and the processor 602.

[0138] The memory 601 is used to store a computer program executable on the processor 602.

[0139] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0140] If the memory 601, the processor 602, and the communication interface 603 are independently implemented, the communication interface 603, the memory 601, and the processor 602 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 6It is represented by only one thick line, but it does not mean that there is only one bus or one type of bus.

[0141] Optionally, in a specific implementation, if the memory 601, the processor 602, and the communication interface 603 are integrated on a single chip, the memory 601, the processor 602, and the communication interface 603 can communicate with each other through an internal interface.

[0142] The processor 602 may be a central processing unit (CPU for short), or an application specific integrated circuit (ASIC for short), or one or more integrated circuits configured to implement the embodiments of the present application.

[0143] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the calibration method of the vehicle-mounted camera as described above.

[0144] The embodiments of the present application also provide a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the calibration method of the vehicle-mounted camera provided by the embodiments of the present invention.

[0145] In the description of this specification, the descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without conflict, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0146] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0147] Any process or method description represented in a flowchart or otherwise described herein can be understood to represent a module, segment, or portion of code including one or N executable instructions for implementing a customized logical function or process. The scope of the preferred embodiments of the present application includes additional implementations, where functions may be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of the present application pertain.

[0148] The logic and / or steps represented in a flowchart or otherwise described herein, for example, can be considered a sequenced list of executable instructions for implementing a logical function, and can be embodied specifically in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of the computer-readable medium include the following: an electrical connection portion (electronic device) having one or N wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable medium on which the program can be printed, as the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0149] It should be understood that various parts of the present application can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0150] Those of ordinary skill in the art can understand that all or part of the steps carried out in implementing the above-described embodiment methods can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0151] In addition, each functional unit in various embodiments of the present application can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0152] The above-mentioned storage medium can be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for calibrating a vehicle-mounted camera, characterized in that: The following steps are involved: Acquire environmental images collected by multiple on-board cameras installed on the vehicle; Stitching the environment images captured by the multiple vehicle-mounted cameras to generate a current panoramic image of the vehicle; At least one abnormal feature is extracted from the current panoramic image of the vehicle, and the vehicle-mounted camera with abnormal setting is calibrated according to the at least one abnormal feature.

2. The method according to claim 1, characterized in that: The extracting at least one abnormal feature from the current panoramic image of the vehicle comprises: Extracting at least one connection area from the current panoramic image; Determine a plurality of stitching features according to a plurality of stitching images corresponding to each connection area, and determine whether a preset connection condition is met based on the plurality of stitching features; If the preset connection condition is met, an abnormal feature is generated based on the multiple splicing features.

3. The method according to claim 1, characterized in that The stitching of the environment images captured by the multiple vehicle-mounted cameras to generate a current panoramic image of the vehicle includes: Performing image preprocessing on the environment images collected by each vehicle-mounted camera respectively to obtain multiple basic images that meet preset quality conditions; Extracting corresponding image features from each basic image based on preset feature requirements, so as to perform image registration using the image features and obtain corresponding registration results; Image fusion is performed based on the registration result to obtain the current panoramic image.

4. The method according to claim 3, characterized in that: The extracting corresponding image features from each basic image based on preset feature requirements includes: Determining the type of feature to be extracted based on the preset feature requirements; Matching a corresponding feature detector using the feature type to be extracted, and setting a corresponding feature descriptor based on the feature detector; The image features are extracted from each of the base images using the feature detector and the feature descriptor.

5. The method according to claim 1, characterized in that Also includes: matching an adjustment value of the vehicle-mounted camera having the abnormal setting according to the at least one abnormal feature; The abnormally-set vehicle camera is adjusted according to the adjustment value of the abnormally-set vehicle camera until the at least one abnormal feature is eliminated.

6. A calibration device for a vehicle-mounted camera, characterized in that: include: An acquisition module is used to acquire environmental images collected by multiple on-board cameras installed on the vehicle; A stitching module, used for stitching the environment images captured by the multiple vehicle-mounted cameras to generate a current panoramic image of the vehicle; The calibration module is used to extract at least one abnormal feature from the current panoramic image of the vehicle, and calibrate the abnormal vehicle-mounted camera according to the at least one abnormal feature.

7. The device according to claim 6, characterized in that The calibration module comprises: An extraction unit, configured to extract at least one connection area from the current panoramic image; A judging unit, configured to determine a plurality of stitching features according to a plurality of stitching images corresponding to each joint area, and judge whether a preset joint condition is satisfied based on the plurality of stitching features; A generating unit is used to generate an abnormal feature based on the multiple splicing features when the preset connection condition is met.

8. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle camera calibration method according to any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the vehicle camera calibration method as described in any one of claims 1 to 5.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed, the calibration method of the vehicle-mounted camera as described in any one of claims 1 to 5 is implemented.

Images