Vehicle damage assessment reminder method, device, electronic device and storage medium

Abstract

The present invention provides a vehicle damage assessment reminder method, device, electronic device and storage medium, which relate to the field of vehicle insurance damage assessment technology. The method includes the steps of: judging whether a component is damaged based on an actual image of the component and treating the damaged component as a key damaged object; obtaining components adjacent to the position of the key damaged object and treating them as possible damaged objects; attempting to identify an actual image containing a possible damaged object in all actual images that have been taken and using it as a target image; if the actual image containing the possible damaged object cannot be identified, generating a first reminder message; if the actual image containing the possible damaged object is successfully identified and a second reminder message is generated when it is uncertain whether the possible damaged object is damaged. The vehicle damage assessment reminder method of the present invention solves the problem that traditional damage assessment methods are extremely dependent on human experience, thereby reducing the impact of human factors and improving the comprehensiveness and accuracy of the damage assessment process.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle insurance damage assessment, and in particular to a vehicle damage assessment reminder method, device, electronic device and storage medium. Background Art

[0002] In existing vehicle damage assessment schemes, vehicle damage assessments generally require the assessor to photograph damaged parts for evidence collection and then provide a corresponding damage assessment price. However, this damage assessment method is extremely dependent on the assessor's professional experience. If the assessor is inexperienced, he or she may often only observe the obviously damaged parts and ignore or misjudge the damage to other related parts, resulting in the omission of photos of some damaged parts during the shooting process, which in turn causes trouble for subsequent claims.

[0003] There is currently no effective technical solution to the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a vehicle damage assessment reminder method, device, electronic device and storage medium, which solves the problem that traditional damage assessment methods are extremely dependent on human experience, thereby reducing the impact of human factors and improving the comprehensiveness and accuracy of the damage assessment process.

[0005] In a first aspect, the present invention provides a vehicle damage assessment reminder method, comprising the following steps:

[0006] S1. Determine whether the component is damaged based on the actual image of the component, and identify the damaged component as a key damaged object;

[0007] S2. According to a preset database, obtain components adjacent to the location of the key damaged object and identify them as potentially damaged objects;

[0008] S3. Attempting to identify, from all the captured actual images, an actual image containing the potentially damaged object and use it as a target image, specifically comprising:

[0009] S31. After attempting to identify the actual image containing the potentially damaged object, execute:

[0010] S311. If the actual image containing the potentially damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the potentially damaged object;

[0011] S312. If the actual image containing the potentially damaged object is successfully identified, whether the potentially damaged object is damaged is determined based on the target image, and when it is uncertain whether the potentially damaged object is damaged, a second reminder message is generated to remind the user to review the target image, and when it is determined that the potentially damaged object is damaged, the potentially damaged object is converted into the key damaged object.

[0012] The vehicle damage assessment reminder method provided by the present invention provides reminders to the damage assessor when necessary based on the positional correlation between parts and combined with the recognition and analysis of images, so that the damage assessor can conduct damage assessment in a more unified and standardized manner, which is conducive to reducing the situation where the damage assessor misjudgments and omissions in evidence collection are achieved, thereby achieving the effect of improving the comprehensiveness and accuracy of the damage assessment process.

[0013] Furthermore, the specific steps in step S312 include:

[0014] S3121. Potentially damaged objects that are uncertain whether they are damaged and are adjacent to multiple key damaged objects are taken as key inspection objects, and when the key inspection objects appear, additional mandatory requirement information is generated; the mandatory requirement information includes information requiring the user to take actual images of the key inspection objects at a specified angle.

[0015] Key inspection objects are more likely to be damaged and should be given special attention and inspected comprehensively. The purpose of this is to obtain more comprehensive and clearer image information so as to more accurately determine whether these parts are damaged.

[0016] Furthermore, the mandatory requirement information includes information requiring the user to capture six basic views of the key inspection object as actual images of the key inspection object.

[0017] By requiring six basic views, this approach effectively addresses the issues of insufficient information and incomplete perspectives that can arise in traditional damage assessment methods. Appraisers can use this comprehensive image data to more accurately assess the damage to key objects, reducing misjudgments and omissions caused by limited viewing angles.

[0018] Furthermore, the specific steps in step S3 also include:

[0019] S32. Before step S31, the following steps are also included:

[0020] S321. Grouping all the actual images captured according to the assembly to which each component belongs, so that all actual images corresponding to all components belonging to the same assembly are included in the same atlas;

[0021] S322. According to the assembly to which the possibly damaged object belongs, attempt to identify an actual image containing the possibly damaged object in the corresponding atlas.

[0022] By grouping images first, a more precise search range is provided for the subsequent recognition process, significantly reducing the number of images that need to be processed. This not only improves recognition speed but also reduces the probability of misidentification.

[0023] Furthermore, the specific steps in step S1 include:

[0024] S11. Classify all damaged parts into slightly damaged and severely damaged categories according to the degree of damage of each damaged part, and identify the severely damaged parts as the key damaged objects.

[0025] Furthermore, the specific steps in step S11 include:

[0026] S111. Capture six basic views of the damaged vehicle as a damaged image of the damaged vehicle;

[0027] S112. Determine the main damaged position of the damaged vehicle based on the damage image;

[0028] S113. Classify the damaged parts located at the main damaged location into the severely damaged category.

[0029] Furthermore, the specific steps in step S112 include:

[0030] S1121. Obtain the accident type of the vehicle being assessed for damage;

[0031] S1122. Determine the main damaged direction based on the damage image and the accident type.

[0032] In a second aspect, the present invention provides a vehicle damage assessment reminder device, comprising:

[0033] A judgment module is used to judge whether the component is damaged based on the actual image of the component, and to identify the damaged component as a key damaged object;

[0034] An acquisition module, configured to acquire components adjacent to the location of the key damaged object as possible damaged objects based on a preset database;

[0035] The recognition module is configured to attempt to identify, from among all the actual images captured, an actual image containing the potentially damaged object and use it as a target image, specifically comprising:

[0036] S31. After attempting to identify the actual image containing the potentially damaged object, execute:

[0037] S311. If the actual image containing the potentially damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the potentially damaged object;

[0038] S312. If the actual image containing the potentially damaged object is successfully identified, whether the potentially damaged object is damaged is determined based on the target image, and when it is uncertain whether the potentially damaged object is damaged, a second reminder message is generated to remind the user to review the target image, and when it is determined that the potentially damaged object is damaged, the potentially damaged object is converted into the key damaged object.

[0039] The vehicle damage assessment reminder device provided by the present invention assists the damage assessor in completing a more comprehensive and accurate damage assessment through a dual reminder method, thereby greatly reducing the possibility of human error.

[0040] In a third aspect, the present invention provides an electronic device comprising a processor and a memory, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the vehicle damage assessment reminder method provided in the first aspect are executed.

[0041] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps in the vehicle damage assessment reminder method provided in the first aspect are executed.

[0042] From the above, it can be seen that the vehicle damage assessment reminder method provided by the present invention, based on the correlation between parts and combined with image recognition technology, avoids the damage assessor from missing the damage inspection of other related parts, reduces the degree of dependence on the professional experience of the damage assessor in the damage assessment process, and performs damage assessment in a more standardized and objective manner, which is conducive to reducing human misjudgment, omissions and other problems, and achieves the effect of improving the comprehensiveness and accuracy of the damage assessment process.

[0043] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the embodiments of the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flowchart of a vehicle damage assessment reminder method provided in an embodiment of the present invention.

[0045] Figure 2 A schematic structural diagram of a vehicle damage assessment reminder device provided in an embodiment of the present invention.

[0046] Figure 3A schematic structural diagram of an electronic device provided by an embodiment of the present invention.

[0047] Description of labels:

[0048] 100, judgment module; 200, acquisition module; 300, identification module; 13, electronic device; 1301, processor; 1302, memory; 1303, communication bus. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0050] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are used only to distinguish the description and should not be understood as indicating or implying relative importance.

[0051] With the rapid development of the auto insurance business, the accuracy and efficiency of the vehicle damage assessment process have become increasingly important. Traditional vehicle damage assessment methods rely primarily on the adjuster's experience and expertise, assessing the extent of vehicle damage through on-site inspections and photographic documentation. However, this approach has several limitations, including a high reliance on individual experience, the difficulty of identifying certain damaged parts in complex accident scenarios, the risk of omissions caused by time pressures, and the challenges posed by the increasing complexity of modern vehicles.

[0052] For example, a sedan was involved in a side-on collision at an intersection, leaving the right door significantly dented and the right sideview mirror knocked off. Upon arriving at the scene, the damage assessor quickly took photos of the damaged door and sideview mirror. However, due to time constraints and limited experience, he or she may have overlooked potential damage to the right B-pillar and wheel hub. These overlooked areas of damage could lead to an underestimated repair cost and even compromise vehicle safety.

[0053] In this regard, please refer to the attached Figure 1 The present invention provides a vehicle damage assessment reminder method, comprising the following steps:

[0054] S1. Determine whether the component is damaged based on the actual image of the component, and designate the damaged components as key damaged objects. (It should be noted that this embodiment does not limit all damaged components to being key damaged objects. In practice, all damaged components may be designated as key damaged objects, or only a portion may be designated as key damaged objects. For example, in the following embodiment, all damaged components may be classified as lightly damaged or severely damaged, with the severely damaged category belonging to some of the damaged components. In this embodiment, only the severely damaged components are designated as key damaged objects.)

[0055] S2. Based on a preset database, obtain components adjacent to the location of the key damaged object and identify them as potentially damaged objects;

[0056] S3. Among all actual images captured, attempt to identify actual images containing potentially damaged objects and use them as target images, specifically including:

[0057] S31. After attempting to identify an actual image containing a potentially damaged object, perform:

[0058] S311. If the actual image containing the possibly damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the possibly damaged object;

[0059] S312. If the actual image containing the potentially damaged object is successfully identified, whether the potentially damaged object is damaged is determined based on the target image, and when it is uncertain whether the potentially damaged object is damaged, a second reminder message is generated to remind the user to review the target image, and when it is determined that the potentially damaged object is damaged, the potentially damaged object is converted into a key damaged object.

[0060] This embodiment assists the assessor in completing damage assessments through dual reminders. First, if the system detects that the assessor has not taken actual images of potentially damaged objects associated with a key damaged object, the assessor is given a first reminder to correct the error, thereby facilitating a comprehensive damage assessment. Furthermore, if the assessor's actual images of potentially damaged objects fail to confirm damage, a second reminder is given to the assessor to re-examine the potentially damaged objects, thereby facilitating an accurate damage assessment.

[0061] Specifically, the mobile device's camera first captures an image of the damaged vehicle. After capturing the image, the app uses a pre-trained deep learning model, such as a convolutional neural network based on the ResNet-50 architecture, to analyze the image. This model can identify vehicle parts in the image and determine whether they are damaged.

[0062] When the system identifies a damaged component, it marks it as a key damaged object. It then queries a pre-set database that stores the positional relationships between vehicle components. For example, if the front bumper is identified as damaged, the system automatically identifies adjacent components such as the headlights, radiator grille, and front fenders and marks them as potentially damaged.

[0063] The system then searches all previously captured images for these potentially damaged objects. If images of certain potentially damaged objects are missing or unclear, the system generates a reminder message, prompting the user to take additional or new images. For potentially damaged objects for which images are already available, the system uses a deep learning model to analyze whether they are damaged. If damage is uncertain, the system generates a reminder message, prompting the user to recheck.

[0064] This method, through systematic analysis and reminder mechanisms, effectively reduces omissions caused by human error and improves the accuracy and comprehensiveness of vehicle damage assessments. It not only helps inexperienced adjusters avoid missing important information, but also provides assistance to experienced adjusters, ensuring the integrity of the damage assessment process.

[0065] The innovation of this invention lies in combining the correlation of adjacent components, image recognition technology, and an intelligent reminder system to form a closed-loop damage assessment process. This method enables a more comprehensive assessment of vehicle damage, significantly reducing the potential for human error. It also improves damage assessment efficiency. Traditional methods may require assessors to make multiple round trips to inspect the vehicle. However, this method, through the intelligent reminder system, can record all damage details as completely as possible in a single inspection, reducing duplication of work and saving time and labor costs.

[0066] Compared to existing technologies, the solution of this invention not only utilizes advanced deep learning technology for damage identification, but also introduces component correlation analysis and intelligent reminder mechanisms. This method can systematically consider potential damage to adjacent components, rather than just identifying obvious damaged areas.

[0067] Furthermore, the solution of the present invention is highly adaptable and scalable. As vehicle technology continues to advance and new models and components emerge, the system of the present invention can adapt to these changes by updating the preset database and deep learning model, maintaining its effectiveness and accuracy.

[0068] The vehicle damage assessment reminder method proposed in the present invention effectively solves the problem of missing damaged parts during vehicle damage assessment by combining advanced image recognition technology, parts correlation analysis and an intelligent reminder system, significantly improving the accuracy, comprehensiveness and efficiency of damage assessment, and providing insurance companies and car owners with more reliable and efficient damage assessment services.

[0069] In some embodiments, the specific steps in step S312 include:

[0070] S3121. Potentially damaged objects that are uncertain whether they are damaged and are adjacent to multiple key damaged objects are treated as key inspection objects, and when key inspection objects appear, additional mandatory requirement information is generated; the mandatory requirement information includes information requiring the user to take actual images of the key inspection objects at a specified angle.

[0071] In order to more effectively identify and inspect potentially damaged vehicle parts, especially those adjacent to multiple confirmed damaged parts, the concept of "key inspection objects" is introduced in this embodiment, and a special processing flow is set up for them. Key inspection objects refer to potentially damaged objects that are uncertain whether they are damaged but are adjacent to multiple confirmed damaged parts (key damaged objects). For these key inspection objects, the system will additionally generate mandatory information. This mandatory information requires the user to take actual images of the key inspection objects at a specified angle. Compared with other adjacent parts, key inspection objects are more likely to be damaged and should be given special attention and a comprehensive inspection. The purpose of this is to obtain more comprehensive and clearer image information in order to more accurately determine whether these parts are damaged.

[0072] There are multiple ways to generate mandatory information. The system can automatically generate optimal shooting angle recommendations based on the location and characteristics of the key inspection objects. For example, exterior vehicle parts may require multiple angles, while interior parts may require specific disassembly and shooting instructions.

[0073] Furthermore, this technical solution can be combined with image recognition technology to automatically determine whether the image provided by the user meets the requirements. If the image quality is insufficient or the angle is inappropriate, the system can provide immediate feedback and ask the user to retake the photo, thus ensuring the acquisition of high-quality image data.

[0074] This method not only improves the accuracy of vehicle damage assessments but also reduces missed detections and misjudgments, thereby enhancing the efficiency and reliability of the entire damage assessment process. By forcing users to shoot at specific angles, it also standardizes user operations to a certain extent, ensuring high-quality image data and providing a more reliable basis for subsequent damage assessment analysis.

[0075] During implementation, the system first identifies key damaged objects and then, based on a pre-set database, locates adjacent components to these key damaged objects. Next, the system analyzes these adjacent components for uncertain damage and marks them as potentially damaged. Within these potentially damaged objects, the system further selects components adjacent to multiple key damaged objects and marks them as key targets for inspection.

[0076] Once the key inspection objects are identified, the system generates mandatory information. This information not only includes instructions for users to photograph the key inspection objects, but also includes specific shooting angle guidance. For example, the system may require users to take a picture from the front, side, and top to ensure that any possible damage is fully captured.

[0077] After the user follows the instructions to take and upload the images, the system analyzes them. If the image quality is insufficient or the angle doesn't meet the requirements, the system prompts the user to retake the image. Only when all the necessary images meet the requirements will the system proceed to the next stage of damage analysis.

[0078] In actual applications, a certain component is adjacent to multiple damaged components (key damaged objects), but no damage is found during the initial inspection. In some cases, it may be because the angle of observation of the component during the initial inspection is not appropriate, and its damaged point is obscured or appears at another angle, resulting in the component being missed. Specifically, suppose that in a front-end collision accident, the system has confirmed that the front bumper and hood are damaged. According to the preset database, the system knows that the headlights are adjacent to the front bumper and the hood. At this time, the headlights are marked as key inspection objects. Compared with other components that are only adjacent to one damaged component, the headlights have a higher probability of damage. However, the headlights were not found to be damaged during the initial inspection. After a detailed inspection, it was found that at the angle of the initial inspection, the ambient light obscured the cracks on the headlights, while the cracks could be clearly observed at another angle.

[0079] When the system identifies key inspection objects, it will generate mandatory information, requiring users to take photos of these key inspection objects from multiple angles. For example, for headlights, the system may require users to take three photos: the front, a 45-degree angle, and the side; for the hood, the system may require users to take a top view and a side view; for the radiator, the system may require users to open the hood and take photos of the radiator from the front and side.

[0080] In this way, the present invention can more comprehensively capture damage that might otherwise go unnoticed. For example, while a headlight may appear undamaged, multi-angle photography may reveal a subtle crack in the lampshade or slight deformation of the internal structure. Similarly, a hood may appear intact, but a closer inspection may reveal minor dents or paint damage. A radiator may appear fine on the outside, but a close-up image may reveal a small leak or fin deformation.

[0081] Compared to existing technologies, traditional vehicle damage assessment methods rely heavily on the assessor's experience, making them prone to omissions and misjudgments. Parts adjacent to visibly damaged components but with less obvious damage themselves can be easily overlooked. This invention, by introducing the concept of key inspection targets and combining it with mandatory photography requirements, can systematically identify and inspect these easily overlooked components.

[0082] Furthermore, with traditional methods, even experienced assessors may miss some hidden damage due to improper shooting angles. By specifying the shooting angle, this invention ensures that each key inspection object is recorded from the optimal perspective, greatly improving the accuracy of damage identification.

[0083] The technical solution of the present invention effectively solves the problems of omissions and misjudgments that may occur in the process of vehicle damage assessment through a systematic and standardized approach, improves the accuracy and comprehensiveness of damage assessment, and provides a more reliable basis for subsequent claims processing.

[0084] In some embodiments, the mandatory requirement information includes information requiring the user to capture six basic views of the key verification object as actual images of the key verification object.

[0085] To address the issue of how to more comprehensively capture image information of key inspection objects to improve the accuracy and efficiency of vehicle damage assessment, this embodiment requires users to capture six basic views of key inspection objects. This ensures comprehensive image information of key inspection objects for most damaged parts. These six basic views may include images from different angles, such as front, back, left, right, top, and bottom, providing more comprehensive visual information of key inspection objects.

[0086] Specifically, the six basic views in the present invention generally include: front view, rear view, left view, right view, top view, and bottom view. This all-round shooting method ensures that every angle of the key inspection object is captured, thereby minimizing the possibility of missing key information.

[0087] In practice, these views can be captured using a mobile device's camera or a professional camera. To improve capture accuracy and consistency, guide lines or angle indicators can be provided on the capture interface to help users accurately position and align the capture angle. For example, a virtual cube frame can be displayed on the capture interface to guide users in capturing the key inspection object from six directions.

[0088] Furthermore, the present invention can be combined with image recognition technology to automatically detect whether the captured image meets the requirements of the six basic views. If the image of a certain perspective is unclear or incomplete, the system can prompt the user to retake the image to ensure high-quality image data.

[0089] This approach, combined with the previously mentioned step of prioritizing potentially damaged objects adjacent to multiple key damaged objects for inspection, allows for a more comprehensive assessment of potential damage. By capturing these key inspection objects in six views, assessors can more accurately determine the actual damage condition of these components, reducing misjudgments or omissions due to viewing angle limitations.

[0090] During the vehicle damage assessment process, the technical solution of this invention effectively addresses the issues of insufficient information and incomplete angles that can arise in traditional damage assessment methods by requiring the capture of six basic views. This comprehensive image data allows assessors to more accurately assess the damage to key objects, reducing misjudgments or omissions caused by limited viewing angles. Furthermore, this standardized capture requirement improves the efficiency of the damage assessment process, as assessors can review and evaluate these images according to a unified standard.

[0091] For example, suppose the front bumper is identified as a key inspection target in a vehicle collision. Using the method of the present invention, the user is asked to capture six basic views of the front bumper. This includes not only the conventional front view, but also an overhead view taken from below, which may reveal hidden structural damage beneath the bumper. Similarly, a side view may reveal misalignment between the bumper and adjacent panels, while a top view may reveal surface dents or cracks. This comprehensive image acquisition allows adjusters to more accurately assess the extent of damage, including subtle damage that may be overlooked from a single angle.

[0092] Compared to traditional vehicle damage assessment methods, which often rely on the assessor's experience and may focus only on visible damage while overlooking potentially hidden damage, this new method requires capturing six basic views, ensuring a comprehensive inspection of key areas. This not only improves the accuracy of damage assessments but also reduces the risk of duplicate assessments and subsequent disputes due to insufficient information. Furthermore, the standardized imaging process ensures greater consistency and comparability in assessment results across assessors, further enhancing the fairness and reliability of the entire damage assessment process.

[0093] In some embodiments, the specific steps in step S3 further include:

[0094] S32. Before step S31, the following steps are also included:

[0095] S321. Group all actual images captured according to the assembly to which each component belongs, so that all actual images corresponding to all components belonging to the same assembly are included in the same atlas;

[0096] S322. Based on the assembly to which the potentially damaged object belongs, attempt to identify an actual image containing the potentially damaged object in the corresponding atlas.

[0097] To more efficiently identify actual images of potentially damaged objects, this embodiment utilizes two key technical features: grouping actual images by component, and identifying potentially damaged objects within the corresponding component image collection. These features play a significant role in improving the efficiency of identifying actual images of potentially damaged objects.

[0098] Specifically, grouping actual images by assembly can be achieved in a variety of ways. For example, information about components belonging to the same assembly can be pre-stored based on a vehicle structural diagram or a component relationship database. When capturing actual images, each image can be associated with its corresponding assembly through image recognition technology or manual annotation. Another approach is to leverage image metadata to automatically record the shooting position or angle during capture, and then infer the assembly to which the image belongs based on this information.

[0099] When identifying potentially damaged objects within an atlas of corresponding assembly parts, a variety of recognition algorithms can be employed. For example, deep learning-based object detection algorithms, such as YOLO or Faster R-CNN, can be used. These algorithms, specifically trained for vehicle components, can quickly and accurately locate and identify components within images. Furthermore, traditional computer vision techniques, such as edge detection and shape matching, can be combined to improve recognition accuracy.

[0100] By performing grouping first, the subsequent recognition process provides a more precise search range, significantly reducing the number of images to be processed. This not only improves recognition speed but also reduces the probability of misidentification. For example, if a potentially damaged bumper belongs to an exterior vehicle assembly, the system only needs to search within the image collection of the exterior vehicle assembly, without having to process images of the engine compartment or interior.

[0101] In practical applications, this approach can significantly improve the efficiency and accuracy of the vehicle damage assessment process. For example, in an accident involving damage to the front bumper, the system may identify the front bumper as the main damaged object. It will then mark parts adjacent to the front bumper, such as headlights, radiator grilles, etc., as potentially damaged objects. When trying to identify actual images of these potentially damaged objects, the system first groups all captured images by assembly parts. Since the front bumper, headlights, and radiator grilles all belong to the front body assembly, the system will search in the atlas of the front body assembly without wasting time on other irrelevant images.

[0102] This approach not only improves identification efficiency but also helps adjusters conduct a more comprehensive assessment of vehicle damage. By systematically organizing and analyzing images, the risk of missing potentially damaged parts is reduced, thereby improving the accuracy and completeness of damage assessments. For example, even if an adjuster might overlook a small component adjacent to a significantly damaged front bumper (such as a fog lamp or vehicle emblem), this approach can alert them to inspect these potentially overlooked parts.

[0103] This approach also offers excellent scalability and adaptability. As vehicle models are updated and parts change, simply updating the relational database of assemblies and components will adapt to the new model and component configurations. This flexibility enables continuous optimization and improvement of the system to adapt to evolving vehicle technologies and market demands.

[0104] Compared to traditional vehicle damage assessment methods, which rely heavily on the assessor's experience and subjective judgment, these methods are susceptible to individual limitations in ability and attention, potentially leading to overlooking certain damaged parts. The method presented in this paper, through systematic image organization and intelligent recognition, enables a more comprehensive and objective assessment of vehicle damage, reducing human error and omissions. Furthermore, this method improves the efficiency of the assessment process, reduces the assessor's workload, and enables them to focus more on complex situations requiring professional judgment.

[0105] The present invention optimizes the recognition process of actual images of potentially damaged objects by rationally organizing and utilizing existing image information. This not only improves the efficiency and accuracy of vehicle damage assessment, but also provides a more reliable basis for the insurance claims process, helping to improve the quality of the entire vehicle insurance service.

[0106] In some embodiments, the specific steps in step S1 include:

[0107] S11. All damaged parts are classified into slightly damaged and severely damaged categories according to the degree of damage to each part, and the severely damaged parts are designated as the most damaged.

[0108] To address the issue of how to classify damaged parts based on their severity for more targeted subsequent processing, this embodiment categorizes damaged parts into two categories: minor and major. In practice, severely damaged parts are more likely to have collateral damage to adjacent parts. Therefore, the severely damaged category is prioritized as the damaged part, and subsequent steps are used to investigate the damage to other related parts, ensuring that these parts, which may have caused collateral damage to other parts, receive priority processing. This effectively addresses the issue of how to classify and prioritize damaged parts. This classification method helps assessors more quickly identify parts requiring special attention, thereby improving assessment efficiency. It also provides a foundation for subsequent assessment processes. For example, the location of the major damaged part can be used to further identify potentially damaged adjacent parts, leading to a more comprehensive assessment of the vehicle's damage. Furthermore, this classification method may help improve the accuracy of damage assessments. By distinguishing between minor and major damage, assessors can adopt different assessment strategies for different degrees of damage, resulting in more accurate estimates of repair costs and time. In general, this classification method based on the degree of damage provides a systematic and targeted processing framework for vehicle damage assessment, which helps to improve the efficiency, accuracy and comprehensiveness of damage assessment.

[0109] In practice, the present invention can employ a variety of methods to determine the extent of component damage. For example, image analysis techniques can be used to analyze actual images of components and assess the extent of damage based on factors such as the damaged area, depth, and location. Another approach involves incorporating a pre-defined damage assessment model, which may consider factors such as the component's importance, the difficulty of replacement, and its impact on the vehicle's overall performance.

[0110] The present invention can set a threshold for classifying minor damage from major damage. For example, parts with a damage score exceeding 70 (out of 100) can be classified as major damage, while those with a score of 70 or below can be classified as minor damage. This threshold can be adjusted based on actual conditions to accommodate different vehicle types or insurance policy requirements.

[0111] Combined with the aforementioned embodiments, significant technical benefits can be achieved. For example, when identifying potentially damaged objects, components adjacent to severely damaged components are prioritized, allowing for more targeted inspections and improving the efficiency and accuracy of damage assessments. Furthermore, when generating reminders, the priority and content of reminders can be adjusted based on the severity of component damage, ensuring that assessors pay more attention to severely damaged components.

[0112] In practical application, the method of the present invention can be implemented as follows: First, the damage assessment system uses image recognition technology to analyze the damage to each vehicle component. The system uses a preset scoring model that considers factors such as the damaged area (as a percentage of the total component area), the damage depth (surface scratches, dents, or structural damage), and the component's importance (its impact on vehicle safety and functionality). Each factor is assigned a weight, ultimately resulting in a comprehensive score ranging from 0 to 100.

[0113] For example, the damage to a front bumper is as follows:

[0114] Damaged area: 40% (score: 40 points; weight 0.3);

[0115] Damage depth: deep depression (score: 70 points; weight 0.5);

[0116] Component Importance: Medium (Weight: 1.2);

[0117] Comprehensive score calculation: (40 * 0.3 + 70 * 0.5) * 1.2 = 54 points

[0118] If the system sets the severe damage threshold at 60 points, the front bumper will be classified as slightly damaged. However, if the inspection finds deformation of the anti-collision beam behind the front bumper, the system will reassess the front bumper and may raise the score to above 60 points, classifying it as severely damaged and making it a key damaged object.

[0119] This damage-based classification method offers significant advantages over traditional damage assessment methods. Traditional methods often rely on the assessor's experience and judgment, which is prone to subjective bias and omissions. However, the present method, through an objective scoring system, can more accurately and consistently assess the damage level of components. Furthermore, by prioritizing severely damaged components, assessors can conduct more targeted, in-depth inspections and assessments, thereby improving the accuracy and efficiency of damage assessments.

[0120] Compared to existing technologies, the method of the present invention not only addresses the issue of inexperienced adjusters but also provides a standardized loss assessment process. This method reduces human error, improves the consistency of loss assessments, and provides insurance companies with more detailed and reliable loss assessment data. Furthermore, by distinguishing between minor and major damage, the method of the present invention can better allocate loss assessment resources, ensuring that critical damage receives timely and adequate attention, thereby improving the efficiency and quality of the entire loss assessment process.

[0121] In some embodiments, the specific steps in step S11 include:

[0122] S111. Capture six basic views of the damaged vehicle as a damaged image of the damaged vehicle;

[0123] S112. Determine the primary damaged location of the vehicle based on the damage image;

[0124] S113. Determine that damaged parts are located in the main damaged direction and are classified as severely damaged.

[0125] In actual applications, the situation where multiple parts are damaged together generally occurs in the area of ​​the vehicle that is most severely damaged. This embodiment more accurately determines the main damaged direction of the vehicle, thereby more effectively identifying severely damaged parts. This embodiment improves the accuracy of determining the main damaged direction by systematically collecting and analyzing vehicle damage information. By taking six basic views of the damaged vehicle as damage images, the damage to the vehicle can be fully captured. Judging the main damaged direction based on these damage images can more accurately locate the severely damaged area. Classifying the damaged parts located in the main damaged direction as severely damaged will help to more comprehensively and efficiently find all damaged parts in the area.

[0126] In practice, the six basic views typically include the front, rear, left, right, top, and bottom views of the vehicle. These views can be captured using a high-definition camera or specialized vehicle inspection equipment. Each view should clearly demonstrate the vehicle's overall condition and details.

[0127] The process of determining the primary damage location can combine image recognition technology and expert systems. For example, deep learning algorithms can be used to analyze damage images to identify the location, area, and severity of the damaged area. Simultaneously, a rule-based expert system can be established to infer the most likely primary damage location based on different accident types (such as rear-end collisions and side impacts) and damage characteristics.

[0128] When categorizing parts as severely damaged, specific criteria can be established. For example, a comprehensive assessment can be conducted based on factors such as the location of the primary damage (center or edge), the percentage of damaged area (e.g., exceeding 50%), and the extent of damage (e.g., severity of deformation). Furthermore, the importance and safety of the parts can be considered. For example, critical structural parts or safety system components may be prioritized for severe damage.

[0129] For example, consider a sedan involved in a side collision. First, a camera captures six basic views of the vehicle. While analyzing these images, the image recognition system detects significant dents and deformations in the right side view, covering approximately 40% of the right side of the vehicle. The system also detects that the right front and right rear doors cannot open properly. Based on this information, the system determines that the primary damage is to the right side of the vehicle.

[0130] The system can then classify all damaged parts on the right side as severely damaged, or it can classify all damaged parts on the right side as severely damaged. For example, the right front and right rear doors are severely damaged because they are severely deformed and no longer function properly. The right fender only has surface scratches and is not currently classified as severely damaged. Although the right B-pillar has less obvious damage, given its importance to the vehicle's structural strength, the system recommends classifying it as severely damaged and prompts for further structural inspection.

[0131] Compared to traditional vehicle damage assessment methods, which often rely on the assessor's subjective judgment and are easily limited by personal experience and observation angles, this method may overlook subtle but important damaged areas. This invention, through systematic six-view imaging and intelligent analysis, significantly improves the comprehensiveness and accuracy of damage assessment. In particular, when determining the primary location of damage, the combination of multi-angle visual information and intelligent analysis systems enables more objective and precise identification of severely damaged areas.

[0132] Furthermore, the method of the present invention provides a more targeted inspection and repair strategy by classifying components with significant damage in the primary location as severely damaged. This not only helps improve repair efficiency and quality, but also reduces the risk of missed inspections and misjudgments. For example, in the aforementioned side collision example, even if the B-pillar's exterior damage is minimal, the system will recommend classifying it as severely damaged. This approach effectively prevents safety hazards caused by overlooking potential damage to critical structural components.

[0133] The proposed method significantly improves the accuracy and efficiency of vehicle damage assessment by combining comprehensive visual information collection with intelligent analysis and processing. It not only more precisely determines the primary location of damage but also more effectively identifies severely damaged parts, providing a more reliable basis for subsequent repairs and claims processing, ultimately improving the quality and efficiency of the entire vehicle damage assessment and repair process.

[0134] In some embodiments, the specific steps in step S112 include:

[0135] S1121. Obtain the accident type of the vehicle being assessed for damage;

[0136] S1122. Determine the main damaged location based on the damage image and accident type.

[0137] To more accurately determine the primary location of damage in a vehicle, thereby improving the accuracy and efficiency of vehicle damage assessment, this embodiment involves two key technical features: obtaining the accident type of the vehicle being assessed, and determining the primary location of damage based on the damage image and accident type. Obtaining the accident type provides important context for determining the primary location of damage, and combining the damage image and accident type to determine the primary location of damage enables a more comprehensive and accurate assessment of vehicle damage.

[0138] In practice, obtaining the accident type of a vehicle being assessed for damage can be accomplished in a variety of ways. For example, the accident type can be inferred through manual user input, automatic extraction from insurance company accident reports, or intelligent analysis of vehicle damage images. Accident types may include, but are not limited to, rear-end collisions, side collisions, head-on collisions, and rollovers.

[0139] To determine the primary damage location based on the damage image and accident type, the system can employ machine learning algorithms, such as convolutional neural networks (CNNs), to analyze the damage image. Furthermore, the system combines the accident type information with a pre-trained model to predict the most likely damage location. For example, in a rear-end collision, the system will focus on analyzing damage to the front and rear of the vehicle; whereas, in a side collision, the system will focus more on damage to the side of the vehicle.

[0140] This approach, combined with the aforementioned practice of capturing six basic views of the vehicle, yields significant synergy. The six basic views provide comprehensive damage information, while the inclusion of the accident type provides crucial clues for interpreting these images. This combination not only improves the accuracy of determining the location of primary damage but also helps identify minor damage that might otherwise be overlooked.

[0141] In practice, the system might use a probabilistic model to comprehensively consider the damage image and accident type. For example, in a side impact accident, if the damage image shows significant damage to the right side of the vehicle, the system might assign an 80% probability of damage to the right side. It also considers possible minor damage in other areas, such as a 20% probability of damage to the front. This probabilistic approach provides a more comprehensive picture of vehicle damage and aids in subsequent damage assessment.

[0142] This method allows the system to more accurately identify the primary damage locations on the vehicle, helping to more accurately identify severely damaged parts. This not only improves the accuracy of damage assessments but also increases their efficiency, as the system can more specifically focus on the parts with the most damage. Furthermore, this method may help identify subtle damage. For example, certain accident types may result in seemingly minor but actually severe internal damage. By combining the accident type and damage image, the system may prompt inspection of these subtle areas.

[0143] For example, suppose a vehicle is involved in a side collision. First, the system automatically obtains the accident type as "side collision" from the insurance company's accident report. The system then analyzes the damage images from the six basic views captured. Using a pre-trained CNN model, the system identifies significant dents and scratches on the right door and right fender. Combined with the "side collision" accident type information, the system further infers a 90% probability that the right side is the primary damage area. At the same time, the system also notices minor scratches on the front bumper. Considering that a side collision may cause a slight deflection of the vehicle, the system lists the front as a secondary damage area with a 30% probability.

[0144] Based on this information, the system automatically categorizes right-side components, such as the right door and right fender, as severely damaged, prompting the assessor to focus on inspecting these parts. The system also recommends inspecting the front bumper and related connecting structures to ensure no potential damage has been missed.

[0145] Compared with existing technologies, traditional damage assessment methods rely primarily on the adjuster's experience, which can lead to subjective biases and omissions of subtle damage. However, the present invention combines accident type and damage images, utilizing machine learning algorithms for objective analysis. This not only more accurately determines the location of primary damage but also identifies minor damage that may have been overlooked. This method significantly reduces human error and improves the accuracy and comprehensiveness of damage assessments. Furthermore, due to the system's ability to rapidly process and analyze information, the present invention significantly improves damage assessment efficiency compared to traditional methods. This not only helps insurance companies more accurately assess compensation amounts but also provides car owners with faster and fairer claims services.

[0146] Please refer to Figure 2 , Figure 2 In some embodiments of the present invention, a vehicle damage assessment reminder device is provided. The vehicle damage assessment reminder device is integrated into a back-end control device in the form of a computer program and includes:

[0147] The judgment module 100 is used to judge whether the component is damaged based on the actual image of the component, and to identify the damaged component as a key damaged object;

[0148] The acquisition module 200 is used to acquire components adjacent to the location of the key damaged object according to a preset database and use them as possible damaged objects;

[0149] The recognition module 300 is configured to attempt to identify, from all captured actual images, actual images containing potentially damaged objects and use them as target images, specifically including:

[0150] S31. After attempting to identify an actual image containing a potentially damaged object, perform:

[0151] S311. If the actual image containing the possibly damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the possibly damaged object;

[0152] S312. If the actual image containing the potentially damaged object is successfully identified, whether the potentially damaged object is damaged is determined based on the target image, and when it is uncertain whether the potentially damaged object is damaged, a second reminder message is generated to remind the user to review the target image, and when it is determined that the potentially damaged object is damaged, the potentially damaged object is converted into a key damaged object.

[0153] In some embodiments, the recognition module 300 attempts to identify an actual image containing a potentially damaged object from among all captured actual images and uses the image as a target image. If the actual image containing the potentially damaged object is successfully identified, the recognition module 300 performs the following when it is uncertain whether the potentially damaged object is damaged:

[0154] S3121. Potentially damaged objects that are uncertain whether they are damaged and are adjacent to multiple key damaged objects are treated as key inspection objects, and when key inspection objects appear, additional mandatory requirement information is generated; the mandatory requirement information includes information requiring the user to take actual images of the key inspection objects at a specified angle.

[0155] In some embodiments, the mandatory requirement information includes information requiring the user to capture six basic views of the key verification object as actual images of the key verification object.

[0156] In some embodiments, the recognition module 300 performs the following when attempting to identify an actual image containing a potentially damaged object from among all actual images that have been captured and using it as a target image:

[0157] S32. Before step S31, the following steps are also included:

[0158] S321. Group all actual images captured according to the assembly to which each component belongs, so that all actual images corresponding to all components belonging to the same assembly are included in the same atlas;

[0159] S322. Based on the assembly to which the potentially damaged object belongs, attempt to identify an actual image containing the potentially damaged object in the corresponding atlas.

[0160] In some embodiments, the determination module 100 is used to determine whether a component is damaged based on an actual image of the component and to identify the damaged component as a key damaged object when performing the following operations:

[0161] S11. All damaged parts are classified into slightly damaged and severely damaged categories according to the degree of damage to each part, and the severely damaged parts are designated as the most damaged.

[0162] In some embodiments, the judgment module 100 is configured to classify all damaged components into slightly damaged and severely damaged categories according to the degree of damage of each damaged component, and to prioritize severely damaged components as:

[0163] S111. Capture six basic views of the damaged vehicle as a damaged image of the damaged vehicle;

[0164] S112. Determine the primary damaged location of the vehicle based on the damage image;

[0165] S113. Determine that damaged parts are located in the main damaged direction and are classified as severely damaged.

[0166] In some embodiments, when determining the main damaged position of the damaged vehicle based on the damage image, the judgment module 100 performs the following:

[0167] S1121. Obtain the accident type of the vehicle being assessed for damage;

[0168] S1122. Determine the main damaged location based on the damage image and accident type.

[0169] Please refer to Figure 3 , Figure 3This is a structural diagram of an electronic device provided by an embodiment of the present invention. The present invention provides an electronic device 13, including: a processor 1301 and a memory 1302. The processor 1301 and the memory 1302 are interconnected and communicate with each other via a communication bus 1303 and / or other forms of connection mechanisms (not shown). The memory 1302 stores computer-readable instructions executable by the processor 1301. When the electronic device is running, the processor 1301 executes the computer-readable instructions to execute the vehicle damage assessment reminder method in any optional implementation of the above embodiment to achieve the following functions: determine whether the component is damaged based on the actual image of the component, and determine that the damaged component is a key damaged object; obtain information based on a preset database. Taking parts adjacent to the position of the key damaged object and taking them as possible damaged objects; among all actual images that have been taken, attempting to identify an actual image containing the possible damaged object and taking it as a target image, specifically including: after attempting to identify the actual image containing the possible damaged object, executing: if the actual image containing the possible damaged object cannot be identified, generating a first reminder message for reminding the user to take an actual image of the possible damaged object; if the actual image containing the possible damaged object is successfully identified, judging whether the possible damaged object is damaged based on the target image, and generating a second reminder message for reminding the user to review the target image when it is uncertain whether the possible damaged object is damaged; and converting the possible damaged object into a key damaged object when it is determined that the possible damaged object is damaged.

[0170] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the vehicle damage assessment reminder method in any optional implementation of the above-mentioned embodiment is executed to achieve the following functions: judging whether a component is damaged based on an actual image of the component, and treating the damaged component as a key damaged object; obtaining components adjacent to the position of the key damaged object based on a preset database and treating them as possible damaged objects; attempting to identify an actual image containing a possible damaged object among all actual images that have been taken and using it as a target image, specifically including: after attempting to identify an actual image containing a possible damaged object, executing: if the actual image containing the possible damaged object cannot be identified, generating a first reminder message for reminding the user to take a photo of the actual image of the possible damaged object; if the actual image containing the possible damaged object is successfully identified, judging whether the possible damaged object is damaged based on the target image, and generating a second reminder message for reminding the user to review the target image when it is uncertain whether the possible damaged object is damaged, and converting the possible damaged object into a key damaged object when it is determined that the possible damaged object is damaged.

[0171] Among them, the computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0172] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, the indirect coupling or communication connection of the device or unit may be electrical, mechanical or other forms.

[0173] In addition, the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0174] Furthermore, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0175] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0176] The foregoing description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A vehicle damage assessment reminder method, characterized in that: The following steps are involved: S1. Determine whether the component is damaged based on the actual image of the component, and identify the damaged component as a key damaged object; S2. According to a preset database, obtain components adjacent to the location of the key damaged object and identify them as potentially damaged objects; S3. Attempting to identify, from all the captured actual images, an actual image containing the potentially damaged object and use it as a target image, specifically comprising: S31. After attempting to identify the actual image containing the potentially damaged object, execute: S311. If the actual image containing the potentially damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the potentially damaged object; S312. If the actual image containing the potentially damaged object is successfully identified, determining whether the potentially damaged object is damaged based on the target image. If it is uncertain whether the potentially damaged object is damaged, generating a second reminder message to remind the user to review the target image. If it is determined that the potentially damaged object is damaged, the potentially damaged object is designated as the key damaged object. The specific steps in step S1 include: S11. Classify all confirmed damaged parts into lightly damaged and heavily damaged categories based on the degree of damage to each component, and designate the heavily damaged parts as the key damaged components; The specific steps in step S11 include: S111. Capture six basic views of the damaged vehicle as a damaged image of the damaged vehicle; S112. Determine the main damaged position of the damaged vehicle based on the damage image; S113. Classify the damaged parts located in the main damaged position as severely damaged; The specific steps in step S112 include: S1121. Obtain the accident type of the vehicle being assessed for damage; S1122. Determine the main damaged location based on the damage image and the accident type; The specific steps in step S312 include: S3121. Potentially damaged objects that are uncertain whether they are damaged and are adjacent to multiple key damaged objects are taken as key inspection objects, and when the key inspection objects appear, additional mandatory requirement information is generated; the mandatory requirement information includes information requiring the user to take actual images of the key inspection objects at a specified angle.

2. The vehicle damage assessment reminder method according to claim 1, characterized in that: The mandatory requirement information includes information requiring the user to capture six basic views of the key inspection object as actual images of the key inspection object.

3. The vehicle damage assessment reminder method according to claim 1, characterized in that: The specific steps in step S3 also include: S32. Before step S31, the following steps are also included: S321. Grouping all the actual images captured according to the assembly to which each component belongs, so that all actual images corresponding to all components belonging to the same assembly are included in the same atlas; S322. According to the assembly to which the possibly damaged object belongs, attempt to identify an actual image containing the possibly damaged object in the corresponding atlas.

4. A vehicle damage assessment reminder device, characterized in that: include: A judgment module is used to judge whether the component is damaged based on the actual image of the component, and to identify the damaged component as a key damaged object; An acquisition module, configured to acquire components adjacent to the location of the key damaged object as possible damaged objects based on a preset database; The recognition module is configured to attempt to identify, from among all the actual images that have been taken, an actual image containing the potentially damaged object and use it as a target image, specifically comprising: S31. After attempting to identify the actual image containing the potentially damaged object, execute: S311. If the actual image containing the potentially damaged object cannot be identified, a first reminder message is generated to remind the user to take an actual image of the potentially damaged object; S312. If the actual image containing the potentially damaged object is successfully identified, determining whether the potentially damaged object is damaged based on the target image. If it is uncertain whether the potentially damaged object is damaged, generating a second reminder message to remind the user to review the target image. If it is determined that the potentially damaged object is damaged, the potentially damaged object is designated as the key damaged object. The judgment module is used to determine whether a component is damaged based on the actual image of the component and to identify the damaged component as a key damaged object. S11. Classify all confirmed damaged parts into lightly damaged and heavily damaged categories based on the severity of damage to each part, with heavily damaged parts designated as key damaged parts. The judgment module is used to classify all the damaged parts into lightly damaged and severely damaged categories according to the damage degree of each damaged part, and to classify the severely damaged parts as the key damaged parts. S111. Capture six basic views of the damaged vehicle as a damaged image of the damaged vehicle; S112. Determine the main damaged position of the vehicle based on the damage image; S113. Classify damaged parts located in the main damage direction as severely damaged; The judgment module is used to determine the main damaged position of the damaged vehicle based on the damage image and executes the following: S1121. Obtain the accident type of the vehicle being assessed for damage; S1122. Determine the main damaged location based on the damage image and accident type; The recognition module is used to try to identify the actual image containing the potentially damaged object from all the actual images that have been taken and use it as the target image. If the actual image containing the potentially damaged object is successfully identified and it is determined whether the potentially damaged object is damaged, the following steps are performed: S3121. Potentially damaged objects that are uncertain whether they are damaged and are adjacent to multiple key damaged objects are taken as key inspection objects, and when the key inspection objects appear, additional mandatory requirement information is generated; the mandatory requirement information includes information requiring the user to take actual images of the key inspection objects at a specified angle.

5. An electronic device, characterized in that: It includes a processor and a memory, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the vehicle damage assessment reminder method as described in any one of claims 1 to 3 are executed.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of the vehicle damage assessment reminder method as described in any one of claims 1 to 3 are executed.

Images