Vehicle loss assessment method and device, computer equipment and storage medium

Abstract

The invention belongs to the technical field of artificial intelligence, and discloses a vehicle loss assessment method and device, computer equipment and a storage medium, and the method comprises the steps: carrying out the intelligent photographing of a current to-be-assessed vehicle according to the prompt information of vehicle loss assessment application software, and obtaining a first to-be-assessed vehicle picture; a watermark adding rule of the first to-be-assessed vehicle picture is preset, and watermark information is added; obtaining a second picture of a vehicle to be subjected to loss assessment stored in the vehicle loss assessment application software and judging whether the picture contains the watermark information or not; performing watermark analysis on the third to-be-subjected-to-damage-assessment vehicle picture containing the watermark information to obtain shooting information of the third to-be-subjected-to-damage-assessment vehicle picture; and presetting a part segmentation model for carrying out damage assessment on the vehicle part, and inputting the third to-be-assessed vehicle picture and the shooting information of the third to-be-assessed vehicle picture into the preset part segmentation model for carrying out damage identification on the vehicle part to obtain a damage assessment result of the vehicle part. According to the invention, the accuracy of vehicle loss assessment is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a vehicle damage assessment method, device, computer equipment and storage medium. Background Art

[0002] With the rapid development of technologies such as artificial intelligence, smartphones have the ability to integrate many lightweight models and apply them in various scenarios, thus providing users with a convenient, intelligent and personalized experience. At present, the relatively mature vehicle damage assessment method in the existing technology is to detect vehicle parts and vehicle damage through computer vision algorithms, detect the damaged vehicle parts in the picture and recommend corresponding repair plans, so as to achieve image-level vehicle damage assessment. However, this method requires the deployment of vehicle parts detection models and vehicle damage detection models in the backend service cluster, which consumes a lot of computing resources. In addition, when assessing vehicle damage on a single image, the degree of damage to the same component between different images will be very different due to problems such as irregular picture shooting and visual angle, which further leads to poor vehicle damage assessment results.

[0003] In view of this, how to improve the accuracy of vehicle damage assessment is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The present invention provides a vehicle damage assessment method, device, computer equipment and storage medium to solve the technical problem that the existing vehicle damage assessment method has poor effect when assessing vehicle damage.

[0005] In a first aspect, the method is applied to a vehicle damage assessment application software, characterized in that a vehicle component detection model and a vehicle damage detection model are integrated in the vehicle damage assessment application software, and the method comprises:

[0006] Intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage;

[0007] Presetting a watermark adding rule for the first image of the vehicle to be assessed for damage, and adding watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule;

[0008] Obtaining a second image of the vehicle to be damaged stored in the vehicle damage assessment application software, and determining whether the second image of the vehicle to be damaged contains watermark information;

[0009] Performing watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to obtain shooting information of the third damaged vehicle picture;

[0010] A component segmentation model for assessing damage to vehicle components is preset, and the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage are input into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

[0011] In a second aspect, a vehicle damage assessment device is provided, the device being used to implement the vehicle damage assessment method as described in the first aspect, comprising:

[0012] A shooting module is used to intelligently shoot the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first picture of the vehicle to be assessed for damage;

[0013] A watermark adding module, used to preset a watermark adding rule for the first image of the vehicle to be assessed for damage, and add watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule;

[0014] A judgment module, used for obtaining a second image of the vehicle to be assessed for damage stored in the vehicle damage assessment application software, and judging whether the second image of the vehicle to be assessed for damage contains watermark information;

[0015] A watermark analysis module, used to perform watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to be assessed, to obtain shooting information of the third damaged vehicle picture to be assessed;

[0016] The recognition module is used to preset a component segmentation model for assessing damage to vehicle components, and input the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

[0017] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned vehicle damage assessment method when executing the computer program.

[0018] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned vehicle damage assessment method are implemented.

[0019] In the above-mentioned scheme implemented based on the vehicle damage assessment method, device, computer equipment and storage medium, the vehicle component detection model and the vehicle damage detection model can be obtained through the client, the obtained vehicle component detection model and the vehicle loss detection model can be integrated into the vehicle damage assessment software development toolkit, and the vehicle damage assessment software development toolkit can be integrated into the vehicle damage assessment application software; the vehicle to be assessed is intelligently photographed according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed; a watermark adding rule for the first image of the vehicle to be assessed is preset, and watermark information is added to the first image of the vehicle to be assessed according to the preset watermark adding rule; the second image of the vehicle to be assessed stored in the vehicle damage assessment application software is obtained, and it is determined whether the second image of the vehicle to be assessed contains watermark information. Watermark information; watermark analysis is performed on the third image of the vehicle to be assessed that contains the watermark information in the second image of the vehicle to be assessed, so as to obtain shooting information of the third image of the vehicle to be assessed; the third image of the vehicle to be assessed and the shooting information of the third image of the vehicle to be assessed are input into the component segmentation model to identify the damage of the vehicle components, so as to obtain the damage assessment result of the vehicle components. In the present invention, by judging whether the image of the vehicle to be assessed in the vehicle damage assessment software contains watermark information, the image of the vehicle to be assessed containing the watermark is analyzed, and the traceable watermark information such as the shooting time and the shooting location of the image of the vehicle to be assessed is obtained, so as to accurately locate the time and space background of the vehicle damage, effectively help the assessor to judge the specific damage condition of the vehicle when it is damaged, thereby effectively improving the accuracy of the vehicle damage assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0021] Figure 1 is a schematic diagram of an application environment of a vehicle damage assessment method in one embodiment of the present invention;

[0022] Figure 2 is a schematic diagram of a process of a vehicle damage assessment method in one embodiment of the present invention;

[0023] Figure 3 yes Figure 2 A schematic flow chart of a specific implementation of step S10;

[0024] Figure 4 yes Figure 2 A schematic flow chart of a specific implementation of step S20;

[0025] Figure 5yes Figure 2 A schematic flow chart of a specific implementation of step S30;

[0026] Figure 6 yes Figure 2 A schematic flow chart of a specific implementation of step S40;

[0027] Figure 7 yes Figure 2 A schematic flow chart of a specific implementation of step S50;

[0028] Figure 8 is a structural schematic diagram of a vehicle damage assessment device in one embodiment of the present invention;

[0029] Fig. 9 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0030] Fig.10 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] The vehicle damage assessment method provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In the application environment, Figure 1It is a schematic diagram of an application environment of a vehicle damage assessment method in one embodiment of the present invention; wherein, the client communicates with the server via a network. The server can obtain a vehicle component detection model and a vehicle damage detection model through the client, integrate the obtained vehicle component detection model and vehicle loss detection model into a vehicle damage assessment software development toolkit, and integrate the vehicle damage assessment software development toolkit into a vehicle damage assessment application software; intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage; preset a watermark adding rule for the first image of the vehicle to be assessed for damage, and add watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule; obtain a second image of the vehicle to be assessed for damage stored in the vehicle damage assessment application software, and determine whether the second image of the vehicle to be assessed for damage contains watermark information; and The third damaged vehicle picture containing watermark information in the film is watermark parsed to obtain the shooting information of the third damaged vehicle picture; the third damaged vehicle picture and the shooting information of the third damaged vehicle picture are input into the component segmentation model to identify the damaged vehicle parts, and the damage assessment result of the vehicle parts is obtained. In the present invention, by judging whether the damaged vehicle picture in the vehicle damage assessment software contains watermark information, the damaged vehicle picture containing watermark is parsed, and the time and space background of the damaged vehicle can be accurately located through the parsed traceable watermark information such as the shooting time and shooting location of the damaged vehicle picture, which effectively helps the damage assessor to judge the specific damage condition when the vehicle is damaged, thereby effectively improving the accuracy of vehicle damage assessment. The present invention is described in detail through specific embodiments below.

[0033] See also Figure 2 As shown, Figure 2 A schematic diagram of a process flow of a vehicle damage assessment method provided by an embodiment of the present invention, wherein the vehicle damage assessment method specifically comprises the following steps:

[0034] S10: Intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage.

[0035] When the present invention is used to assess vehicle damage, the vehicle component detection model and the damage detection model can be integrated into the vehicle damage assessment software development toolkit first, and the calling and management functions of the vehicle component detection model and the damage detection model can be realized in the vehicle damage assessment software development toolkit. Then, the vehicle damage assessment software development toolkit is integrated into the vehicle damage assessment application software, and the customer can use it in the vehicle damage assessment application software to take standardized pictures, so as to ensure the comprehensiveness and accuracy of the photographed vehicle information to be assessed, and facilitate the subsequent claims and repair work. For example, the vehicle damage assessment application software can be an automobile insurance claims software. In the automobile insurance industry, in order to facilitate customers to carry out the vehicle damage assessment and claims process, the insurance company can develop a special insurance claims software. The vehicle component detection model integrated in the automobile insurance claims software can accurately identify various key components of the vehicle, such as license plates, frame numbers, headlights, doors, bumpers and other information. The vehicle damage detection model integrated in the automobile insurance claims software can detect scratches, dents, collisions and other damages on the surface of the vehicle.

[0036] After the vehicle damage assessment software is integrated with the vehicle component detection model and the vehicle damage detection model, the vehicle to be assessed can be intelligently photographed according to the prompt information of the vehicle damage assessment application software. The present invention integrates a voice prompt module in the vehicle damage assessment software to guide the user to intelligently photograph the vehicle to be assessed. Specifically, Figure 3 As stated, Figure 3 yes Figure 2 A specific implementation flow diagram of step S10 in the embodiment of the present invention specifically includes the following steps S11-S14:

[0037] S11: Analyze the video frames in the real-time picture based on the vehicle component detection model integrated in the vehicle damage assessment application software to detect whether the real-time picture contains the license plate number and frame number of the vehicle to be assessed. Specifically, the present invention can first use the camera function of the mobile device to take a picture of the vehicle to be assessed; then analyze the video frames in the real-time picture based on the vehicle component detection model integrated in the vehicle damage assessment application software to detect whether the real-time picture contains the license plate number and frame number of the vehicle to be assessed.

[0038] S12: If the license plate number and frame number of the vehicle to be assessed are detected in the real-time image, a perspective image of the part to be detected of the vehicle to be assessed is taken according to the voice prompt information of the vehicle damage assessment application software. Specifically, if the license plate and frame number of the vehicle to be assessed are detected in the real-time image in step S11, the vehicle damage assessment application software prompts the user through voice "The license plate and frame number have been identified, please take a perspective image of the damaged part".

[0039] S13: Analyze the perspective image of the part to be detected of the vehicle to be damaged based on the vehicle damage detection model integrated in the vehicle damage assessment application software to detect whether the part to be detected of the vehicle to be damaged includes a damaged area. Specifically, after taking the perspective image of the part to be detected of the vehicle to be damaged, the present invention needs to continue to use the damage detection model integrated in the vehicle damage assessment application software to perform a preliminary analysis on the perspective image taken by the user to confirm whether the image contains a possible damaged area.

[0040] S14: If it is detected that the part to be detected of the vehicle to be damaged includes a damaged area, a close-up image of the part to be detected of the vehicle to be damaged is taken according to the voice prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be damaged. Specifically, if a suspected damaged area of ​​the vehicle to be damaged is detected in step S13, the user is prompted by voice to "take a close-up image of the damaged part".

[0041] In addition, when users are taking smart photos, the software development kit in the damaged vehicle application software will provide real-time feedback on the test results to the user. For example, if the license plate or frame number is not clear, the damaged vehicle development software will give a voice prompt "The license plate / frame number is not clear, please take a new photo". For the long-range and close-up photos of the damaged vehicle, if the shooting angle, distance, etc. do not meet the requirements, the damaged vehicle software will give a voice prompt to guide the user to take a new photo to obtain a standardized first damaged vehicle photo.

[0042] S20: Preset a watermark adding rule for the first image of the vehicle to be assessed, and add watermark information to the first image of the vehicle to be assessed according to the preset watermark adding rule. Specifically, in actual use scenarios, the customer needs to forward the first image of the vehicle to be assessed through other software such as WeChat. In order to store and extract information from the shooting stage, we have added a watermark to the processing stage of the vehicle to be assessed in the vehicle damage assessment application software, so that the shooting information of the first image of the vehicle to be assessed can be effectively parsed. Specifically, Figure 4 As stated, Figure 4 yes Figure 2 A specific implementation flow diagram of step S20 in the embodiment of the present invention specifically includes the following steps S21-S22:

[0043] S21: According to the size information of the first damaged vehicle picture to be determined, a white border of a corresponding size is added to the bottom of the first damaged vehicle picture to be determined, and the size information of the first damaged vehicle picture to be determined is updated. Specifically, the present invention first defines a watermark adding rule for the first damaged vehicle picture to be determined, that is, a white border is added to the bottom of the first damaged vehicle picture to be determined according to a corresponding ratio, so that the ratio of the width to the height of the white border is consistent with the ratio of the width to the height of the original first damaged vehicle picture to be determined, and then the size information of the first damaged vehicle picture to be determined after the white border information is added is updated.

[0044] S22: Adding watermark information to the first damaged vehicle picture after the size information is updated according to a preset text content format to obtain the first damaged vehicle picture after the watermark information is added. Specifically, after adding a white border to the first damaged vehicle picture, the present invention adds text content on the rightmost side of the white border, and the format of the text content is "whether smart shooting + component code + whether detailed image", to obtain the first damaged vehicle picture after the watermark information is added.

[0045] S30: Obtain the second image of the vehicle to be assessed stored in the vehicle damage assessment application software, and determine whether the second image of the vehicle to be assessed contains watermark information. Before performing watermark analysis on the second image of the vehicle to be assessed stored in the vehicle damage assessment application software, it is necessary to determine whether the second image of the vehicle to be assessed contains watermark information. Specifically, Figure 5 As stated, Figure 5 yes Figure 2 A specific implementation flow diagram of step S30 in the embodiment of the present invention specifically includes the following steps S41-S44:

[0046] S31: convert the second damaged vehicle image to RGB format to obtain a fourth damaged vehicle image. Specifically, the present invention determines whether the damaged vehicle image contains watermark information, first converts the second damaged vehicle image to RGB format to obtain a fourth damaged vehicle image after RGB format conversion. Since the computer display is based on the RGB color model for image display, if the second damaged vehicle image is not in RGB format, the computer display needs to convert the second damaged vehicle image to RGB format before displaying it, so as to facilitate the processing of each color channel in the damaged vehicle image.

[0047] S32: Preset the height ratio of the white edge in the fourth damaged vehicle picture to be determined, and predict the height range of the white edge of the fourth damaged vehicle picture to be determined based on the size information of the fourth damaged vehicle picture to be determined and the preset height ratio of the white edge in the fourth damaged vehicle picture to be determined. Specifically, the present invention can calculate the theoretical height value of the white edge based on the preset height ratio of the white edge in the fourth damaged vehicle picture to be determined, thereby determining the height range that may contain the watermark. For example, if the preset height ratio of the white edge in the fourth damaged vehicle picture to be determined is 1 / 10 of the original image height, then the height range of the white edge in the bottom area where the watermark is located can be predicted by the total height of the fourth damaged vehicle picture to be determined.

[0048] S33: Based on the predicted height range of the white edge of the fourth image of the vehicle to be determined, a white edge region of the fourth image of the vehicle to be determined is selected by a center cropping algorithm. Specifically, in different operating systems, due to factors such as image rendering mechanisms and device resolution differences, pixel-level errors may occur when adding watermarks. Such errors may cause the actual position and size of the white edge to slightly deviate from the theoretical calculated values. Therefore, it is necessary to further determine the range of the white edge, and the fourth image of the vehicle to be determined may be center cropped to select a determined region of the white edge.

[0049] S34: Determine whether the white edge region of the fourth damaged vehicle image contains only white pixel values ​​based on the image pixel value analysis function. Specifically, the present invention uses the image pixel value analysis function to determine whether the white edge region of the selected fourth damaged vehicle image contains only pixels with values ​​of (255, 255, 255). If the result returned by the image pixel value analysis function indicates that the region contains only pixels with values ​​of 255 in the RGB channels, it can be determined that the region is a white edge region.

[0050] S35: If the white edge region of the fourth image of the damaged vehicle to be determined only contains white pixel values, it is determined that the second image of the damaged vehicle to be determined contains watermark information. Specifically, if the white edge region of the fourth image of the damaged vehicle to be determined is a white edge, it is necessary to save the right half of the image of the white edge region obtained in step S42. If the white edge region of the fourth image of the damaged vehicle to be determined is not a white edge, it means that the image does not contain the watermark we defined, and no further watermark analysis operation is required.

[0051] S40: Perform watermark analysis on the third damaged vehicle image containing watermark information in the second damaged vehicle image to obtain shooting information of the third damaged vehicle image. When it is detected that the second damaged vehicle image stored in the vehicle damage assessment application software contains watermark information, it is necessary to analyze the third damaged vehicle image containing watermark information. Specifically, Figure 6 As stated, Figure 6 yes Figure 2A specific implementation flow diagram of step S40 in the embodiment of the present invention specifically includes the following steps S41-S44:

[0052] S41: Performing text detection on the third damaged vehicle image through the text detection module in the optical character recognition algorithm to obtain the text region of the third damaged vehicle image. Specifically, the present invention uses the text detection module to perform feature extraction on the third damaged vehicle image, and determines the text region of the third damaged vehicle image by extracting feature information such as color changes, edge information, and texture features from the third damaged vehicle image.

[0053] S42: Performing character recognition on the text area of ​​the detected third damaged vehicle to be determined by using the text recognition module in the optical character recognition algorithm to obtain character information of the third damaged vehicle to be determined. Specifically, the present invention uses the text recognition module in the optical character recognition algorithm to match the extracted text area features with the standard features of various characters stored in the model, and determines the most likely corresponding characters in each text area by calculating similarity, probability and other indicators, thereby converting the image content in the text area into specific character information, and finally obtaining the character information contained in the third damaged vehicle to be determined.

[0054] S43: Preset a watermark adding rule for the third image of the vehicle to be assessed for damage, and use a regular expression to determine whether the character information of the third image of the vehicle to be assessed for damage satisfies the preset watermark adding rule. The watermark adding rule for the third image of the vehicle to be assessed for damage is set according to actual needs and business logic. For example, the watermark adding rule for the third image of the vehicle to be assessed for damage can be preset as: "Whether it is intelligent shooting + component code + whether it is a detailed image". Based on the preset watermark adding rule for the third image of the vehicle to be assessed for damage, a regular expression can be used to determine whether the character information of the third image of the vehicle to be assessed for damage satisfies the preset watermark adding rule.

[0055] S44: If the character information of the third image of the damaged vehicle to be assessed meets the preset watermark adding rule, extract the shooting information of the third image of the damaged vehicle to be assessed. When it is determined that the character information of the third image of the damaged vehicle to be assessed meets the preset watermark adding rule, extract the shooting information of the third image of the damaged vehicle to be assessed based on the character information of the third image of the damaged vehicle to be assessed, for example, extract the shooting date, shooting location, shooting equipment number, etc. of the third image of the damaged vehicle to be assessed. Such shooting information can be used for subsequent vehicle damage assessment records, data archiving, tracing, and other related work.

[0056] S50: Preset a component segmentation model for assessing damage to vehicle components, and input the third image of the vehicle to be assessed and the shooting information of the third image of the vehicle to be assessed into the preset component segmentation model to identify damage to vehicle components, and obtain a damage assessment result for the vehicle components. Figure 7As stated, Figure 7 yes Figure 2 A specific implementation flow diagram of step S50 in the embodiment of the present invention specifically includes the following steps S51-S53:

[0057] S51: extracting feature information of the third damaged vehicle image to be determined, and fusing the extracted feature information of the third damaged vehicle image to be determined with the shooting information of the third damaged vehicle image to be determined, to obtain target feature information of the third damaged vehicle image to be determined. Specifically, the shooting information of the third damaged vehicle image to be determined can be used as a priori conditions, and the auxiliary component segmentation model can be used to identify damaged vehicle components, so as to improve the recognition ability of the component segmentation model for the third damaged vehicle image to be determined. Combining the photographed component information with the component segmentation model can more accurately locate the component area and reduce the misdetection of components caused by differences in vehicle models, image quality, etc. For example, if the photographed component is a headlight, then the image should be near the headlight, not the rear of the vehicle. Combining the detail image with the component segmentation model result can reduce missed detection of components. If the left front fender detail is photographed, the component segmentation can only determine that it is a fender due to too many details and cannot determine the orientation. Then, the photographed orientation information can be combined for subsequent judgment.

[0058] S52: Input the target feature information of the fused third image of the vehicle to be damaged into a preset component segmentation model to divide the vehicle components to be damaged. Specifically, the present invention inputs the target feature information of the fused third image of the vehicle to be damaged into the component segmentation model, and the component segmentation model outputs a segmentation map with the same size as the input image. By analyzing the segmentation map, the region where each vehicle component to be damaged is located can be divided.

[0059] S53: Determine the degree of damage of each area of ​​the divided vehicle component to be assessed, and obtain the damage assessment result of the vehicle component. Specifically, the degree of damage of each area of ​​the vehicle component to be assessed can be determined by observing the changes in the appearance characteristics of each area of ​​the divided vehicle component, such as whether there are scratches, dents, etc. on the surface of the component.

[0060] In one embodiment of the present invention, the vehicle damage assessment method further includes: performing data preprocessing on the photographic information of the third image of the vehicle to be assessed, and integrating the photographic information of the third image of the vehicle to be assessed after data preprocessing into the training set of the vehicle component detection model and the vehicle damage detection model. Specifically, the present invention can also reflow the photographic information of the parsed third image of the vehicle to be assessed, iterate the model optimization algorithm, and the detailed image of the vehicle to be assessed can capture minor damage, but often cannot determine the component orientation information. Therefore, the photographic information of the third image of the vehicle to be assessed can be expanded to the training set of the vehicle component detection model and the vehicle damage detection model to enhance the recognition and generalization capabilities of the vehicle damage assessment software for close-up images and minor damage. In addition, the fusion of multimodal information is also considered, and multi-source information such as other sensor data and vehicle historical records is considered to make a comprehensive judgment, and the complementary advantages of these multimodal information are utilized to improve the comprehensive understanding and judgment capabilities of the vehicle damage status.

[0061] It can be seen that in the above scheme, when assessing vehicle damage, by judging whether the image of the vehicle to be assessed in the vehicle damage assessment software contains watermark information, the image of the vehicle to be assessed containing the watermark is analyzed, and the damaged vehicle parts are identified based on the analyzed shooting information, thereby effectively improving the accuracy of vehicle damage assessment.

[0062] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0063] In one embodiment, a vehicle damage assessment device is provided, which corresponds one-to-one to the vehicle damage assessment method in the above embodiment. Figure 8 As shown, Figure 8 8 is a schematic diagram of a structure of a vehicle damage assessment device in one embodiment of the present invention, the vehicle damage assessment device includes a shooting module 81, a watermark adding module 82, a judgment module 83, a watermark parsing module 84, and an identification module 85. The functional modules are described in detail as follows:

[0064] The photographing module 81 is used to intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage;

[0065] A watermark adding module 82, used for presetting a watermark adding rule for the first image of the vehicle to be determined damaged, and adding watermark information to the first image of the vehicle to be determined damaged according to the preset watermark adding rule;

[0066] The judgment module 83 is used to obtain the second image of the vehicle to be assessed for damage stored in the vehicle damage assessment application software, and to judge whether the second image of the vehicle to be assessed for damage contains watermark information;

[0067] A watermark analysis module 84 is used to analyze the watermark of the third damaged vehicle picture containing the watermark information in the second damaged vehicle picture to be determined, and obtain the shooting information of the third damaged vehicle picture to be determined;

[0068] The identification module 85 is used to preset a component segmentation model for assessing damage to vehicle components, and input the third image of the vehicle to be assessed for damage and the shooting information of the third image of the vehicle to be assessed for damage into the preset component segmentation model to identify damage to vehicle components, and obtain a damage assessment result for the vehicle components.

[0069] In one embodiment, the shooting module 81 is specifically used for:

[0070] The vehicle parts detection model integrated in the vehicle damage assessment application software is used to analyze the video frames in the real-time image to detect whether the real-time image contains the license plate number and frame number of the vehicle to be assessed;

[0071] If the real-time image contains the license plate and frame number of the vehicle to be assessed, a long-range image of the part to be detected of the vehicle to be assessed is taken according to the voice prompt information of the vehicle damage assessment application software;

[0072] Based on the vehicle damage detection model integrated in the vehicle damage assessment application software, the long-range image of the part to be inspected of the vehicle to be assessed is analyzed to detect whether the part to be inspected of the vehicle to be assessed contains a damaged area;

[0073] If it is detected that the part to be inspected of the vehicle to be assessed contains a damaged area, a close-up image of the part to be inspected of the vehicle to be assessed is taken according to the voice prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed.

[0074] In one embodiment, the watermark adding module 82 is specifically configured to:

[0075] According to the size information of the first image of the vehicle to be damaged, a white border of a corresponding size is added to the bottom of the first image of the vehicle to be damaged, and the size information of the first image of the vehicle to be damaged is updated;

[0076] The watermark information is added to the first image of the vehicle to be determined damaged after the size information is updated according to a preset text content format to obtain the first image of the vehicle to be determined damaged after the watermark information is added.

[0077] In one embodiment, the determination module 83 is specifically configured to:

[0078] The second image of the vehicle to be damaged is converted into RGB format to obtain a fourth image of the vehicle to be damaged;

[0079] Presetting a height ratio of the white edge in the fourth image of the vehicle to be damaged, and predicting a height range of the white edge in the fourth image of the vehicle to be damaged according to size information of the fourth image of the vehicle to be damaged and the preset height ratio of the white edge in the fourth image of the vehicle to be damaged;

[0080] Based on the predicted height range of the white edge of the fourth image of the vehicle to be damaged, a white edge region of the fourth image of the vehicle to be damaged is selected by a center clipping algorithm;

[0081] Determining whether the white edge area of ​​the fourth to-be-determined damaged vehicle image only contains white pixel values ​​based on the image pixel value analysis function;

[0082] If the white edge region of the fourth image of the damaged vehicle to be assessed only contains white pixel values, it is determined that the second image of the damaged vehicle to be assessed contains watermark information.

[0083] In one embodiment, the watermark parsing module 84 is specifically used to:

[0084] Performing text detection on the third image of the vehicle to be damaged by the text detection module in the optical character recognition algorithm to obtain a text area of ​​the third image of the vehicle to be damaged;

[0085] Performing character recognition on the text area of ​​the detected third vehicle to be damaged by the text recognition module in the optical character recognition algorithm to obtain character information of the image of the third vehicle to be damaged;

[0086] Preset a watermark adding rule for the third image of the vehicle to be assessed for damage, and determine whether the character information of the third image of the vehicle to be assessed for damage satisfies the preset watermark adding rule by using a regular expression;

[0087] If the character information of the third picture of the vehicle to be assessed for damage meets the preset watermark adding rules, the shooting information of the third picture of the vehicle to be assessed for damage is extracted.

[0088] In one embodiment, the identification module 85 is specifically configured to:

[0089] Extracting feature information of the third image of the vehicle to be damaged, and fusing the extracted feature information of the third image of the vehicle to be damaged with the shooting information of the third image of the vehicle to be damaged, to obtain target feature information of the third image of the vehicle to be damaged;

[0090] Inputting the fused target feature information of the third to-be-determined damaged vehicle image into a preset component segmentation model to perform regional division on the to-be-determined damaged vehicle components;

[0091] Determine the degree of damage to each area of ​​the divided vehicle parts to be assessed, and obtain the damage assessment results of the vehicle parts.

[0092] In one embodiment, the vehicle damage assessment device is further used for:

[0093] Data preprocessing is performed on the photographic information of the third image of the vehicle to be damaged, and the photographic information of the third image of the vehicle to be damaged after data preprocessing is integrated into the training set of the vehicle component detection model and the vehicle damage detection model.

[0094] The present invention provides a vehicle damage assessment device. When conducting vehicle damage assessment, by judging whether the image of the vehicle to be assessed in the vehicle damage assessment software contains watermark information, the image of the vehicle to be assessed containing the watermark is analyzed. By analyzing the traceable watermark information such as the shooting time and shooting location of the image of the vehicle to be assessed, the time and space background of the vehicle damage can be accurately located, which effectively helps the assessor to judge the specific damage condition of the vehicle when it is damaged, thereby effectively improving the accuracy of vehicle damage assessment.

[0095] For the specific definition of the vehicle damage assessment device, please refer to the definition of the vehicle damage assessment method above, which will not be repeated here. Each module in the above-mentioned vehicle damage assessment device can be implemented in whole or in part through software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0096] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig. 9 As shown, Fig. 9 It is a structural schematic diagram of a computer device in one embodiment of the present invention. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the service side of a vehicle damage assessment method.

[0097] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Fig.10 As shown, Fig.10: is another structural schematic diagram of a computer device in one embodiment of the present invention. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the client side of a vehicle damage assessment method.

[0098] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0099] Intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage;

[0100] Presetting a watermark adding rule for the first image of the vehicle to be assessed for damage, and adding watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule;

[0101] Obtaining a second image of the vehicle to be damaged stored in the vehicle damage assessment application software, and determining whether the second image of the vehicle to be damaged contains watermark information;

[0102] Performing watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to obtain shooting information of the third damaged vehicle picture;

[0103] A component segmentation model for assessing damage to vehicle components is preset, and the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage are input into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

[0104] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0105] Intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage;

[0106] Presetting a watermark adding rule for the first image of the vehicle to be assessed for damage, and adding watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule;

[0107] Obtaining a second image of the vehicle to be damaged stored in the vehicle damage assessment application software, and determining whether the second image of the vehicle to be damaged contains watermark information;

[0108] Performing watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to obtain shooting information of the third damaged vehicle picture;

[0109] A component segmentation model for assessing damage to vehicle components is preset, and the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage are input into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

[0110] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0111] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0112] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0113] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A vehicle damage assessment method, applied to a vehicle damage assessment application software, characterized in that: The vehicle damage assessment application software integrates a vehicle component detection model and a vehicle damage detection model, and the method includes: Intelligently photograph the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed for damage; Presetting a watermark adding rule for the first image of the vehicle to be assessed for damage, and adding watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule; Obtaining a second image of the vehicle to be damaged stored in the vehicle damage assessment application software, and determining whether the second image of the vehicle to be damaged contains watermark information; Performing watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to obtain shooting information of the third damaged vehicle picture; A component segmentation model for assessing damage to vehicle components is preset, and the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage are input into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

2. The vehicle damage assessment method according to claim 1, characterized in that: The intelligent photographing of the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain the first image of the vehicle to be assessed for damage includes: The vehicle parts detection model integrated in the vehicle damage assessment application software is used to analyze the video frames in the real-time image to detect whether the real-time image contains the license plate number and frame number of the vehicle to be assessed; If the real-time image contains the license plate and frame number of the vehicle to be assessed, a long-range image of the part to be detected of the vehicle to be assessed is taken according to the voice prompt information of the vehicle damage assessment application software; Based on the vehicle damage detection model integrated in the vehicle damage assessment application software, the long-range image of the part to be inspected of the vehicle to be assessed is analyzed to detect whether the part to be inspected of the vehicle to be assessed contains a damaged area; If it is detected that the part to be inspected of the vehicle to be assessed contains a damaged area, a close-up image of the part to be inspected of the vehicle to be assessed is taken according to the voice prompt information of the vehicle damage assessment application software to obtain a first image of the vehicle to be assessed.

3. The vehicle damage assessment method according to claim 1, characterized in that: The step of presetting a watermark adding rule for the first image of the vehicle to be assessed for damage, and adding watermark information to the first image of the vehicle to be assessed for damage according to the pre-set watermark adding rule includes: According to the size information of the first image of the vehicle to be damaged, a white border of a corresponding size is added to the bottom of the first image of the vehicle to be damaged, and the size information of the first image of the vehicle to be damaged is updated; The watermark information is added to the first image of the vehicle to be determined damaged after the size information is updated according to a preset text content format to obtain the first image of the vehicle to be determined damaged after the watermark information is added.

4. The vehicle damage assessment method according to claim 1, characterized in that: The determining whether the second image of the vehicle to be damaged contains watermark information includes: The second image of the vehicle to be damaged is converted into RGB format to obtain a fourth image of the vehicle to be damaged; Presetting a height ratio of the white edge in the fourth image of the vehicle to be damaged, and predicting a height range of the white edge in the fourth image of the vehicle to be damaged according to size information of the fourth image of the vehicle to be damaged and the preset height ratio of the white edge in the fourth image of the vehicle to be damaged; Based on the predicted height range of the white edge of the fourth image of the vehicle to be damaged, a white edge region of the fourth image of the vehicle to be damaged is selected by a center clipping algorithm; Determining whether the white edge area of ​​the fourth to-be-determined damaged vehicle image only contains white pixel values ​​based on the image pixel value analysis function; If the white edge region of the fourth image of the damaged vehicle to be assessed only contains white pixel values, it is determined that the second image of the damaged vehicle to be assessed contains watermark information.

5. The vehicle damage assessment method according to claim 1, characterized in that: The step of performing watermark analysis on the third image of the vehicle to be damaged that contains watermark information in the second image of the vehicle to be damaged to obtain shooting information of the third image of the vehicle to be damaged, comprises: Performing text detection on the third image of the vehicle to be damaged by the text detection module in the optical character recognition algorithm to obtain a text area of ​​the third image of the vehicle to be damaged; Performing character recognition on the text area of ​​the detected third vehicle to be damaged by the text recognition module in the optical character recognition algorithm to obtain character information of the image of the third vehicle to be damaged; Preset a watermark adding rule for the third image of the vehicle to be assessed for damage, and determine whether the character information of the third image of the vehicle to be assessed for damage satisfies the preset watermark adding rule by using a regular expression; If the character information of the third picture of the vehicle to be assessed for damage meets the preset watermark adding rules, the shooting information of the third picture of the vehicle to be assessed for damage is extracted.

6. The vehicle damage assessment method according to claim 1, characterized in that: The step of inputting the third image of the vehicle to be damaged and the shooting information of the third image of the vehicle to be damaged into a preset component segmentation model to identify damage to vehicle components and obtain damage assessment results of vehicle components includes: Extracting feature information of the third image of the vehicle to be damaged, and fusing the extracted feature information of the third image of the vehicle to be damaged with the shooting information of the third image of the vehicle to be damaged, to obtain target feature information of the third image of the vehicle to be damaged; Inputting the fused target feature information of the third damaged vehicle image into a preset component segmentation model to divide the damaged vehicle components into regions; Determine the degree of damage to each area of ​​the divided vehicle parts to be assessed, and obtain the damage assessment results of the vehicle parts.

7. The vehicle damage assessment method according to claim 1, characterized in that: The vehicle damage assessment method further includes: Data preprocessing is performed on the photographic information of the third image of the vehicle to be damaged, and the photographic information of the third image of the vehicle to be damaged after data preprocessing is integrated into the training set of the vehicle component detection model and the vehicle damage detection model.

8. A vehicle damage assessment device, the device being used to implement the vehicle damage assessment method according to any one of claims 1 to 7, characterized in that: include: A shooting module is used to intelligently shoot the vehicle to be assessed for damage according to the prompt information of the vehicle damage assessment application software to obtain a first picture of the vehicle to be assessed for damage; A watermark adding module, used to preset a watermark adding rule for the first image of the vehicle to be assessed for damage, and add watermark information to the first image of the vehicle to be assessed for damage according to the preset watermark adding rule; A judgment module, used for obtaining a second image of the vehicle to be assessed for damage stored in the vehicle damage assessment application software, and judging whether the second image of the vehicle to be assessed for damage contains watermark information; A watermark analysis module, used to perform watermark analysis on the third damaged vehicle picture containing watermark information in the second damaged vehicle picture to be assessed, to obtain shooting information of the third damaged vehicle picture to be assessed; The recognition module is used to preset a component segmentation model for assessing damage to vehicle components, and input the third vehicle image to be assessed for damage and the shooting information of the third vehicle image to be assessed for damage into the preset component segmentation model to identify damage to vehicle components, thereby obtaining a damage assessment result for the vehicle components.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the vehicle damage assessment method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the vehicle damage assessment method as described in any one of claims 1 to 7 are implemented.

Images