Vehicle loss assessment result query method and device, equipment and storage medium

Abstract

The invention relates to the technical field of intelligent decision and the field of financial science and technology, and discloses a vehicle loss assessment result query method, which comprises the following steps that: a vehicle claim settlement operation end uploads vehicle insurance pictures to a background loss assessment processing end in batches, records a picture ID and an uploading request in a request result table during each uploading, and marks the receiving state of each vehicle insurance picture; and the background loss assessment processing end extracts a picture downloading task from the request result table, completes downloading of each car insurance picture, and updates a receiving state. After downloading is completed, the car insurance picture processing tasks are added into a task queue, the car insurance picture processing tasks are extracted from the task queue through a car insurance processing process, and corresponding damage form description is generated. Based on the generated damage form description, a corresponding repair scheme is generated in combination with the repair strategy set, and the repair scheme and the damage result are stored. The processing efficiency of the loss assessment process is improved, the integrity of the query result is ensured, and the claim settlement process is more efficient and smoother.

Description

Technical Field

[0001] The present invention relates to the field of intelligent decision-making technology, and in particular to a vehicle damage assessment result query method, device, equipment and storage medium. Background Art

[0002] In modern society, with the continuous increase in the number of cars, vehicle insurance has become one of the important means to protect the rights and interests of car owners. After a traffic accident, the car owner usually claims through the insurance company. One of the core links in the claims process is vehicle damage assessment. Vehicle damage assessment refers to the assessment of the damage to the damaged vehicle after the traffic accident and the determination of the cost of repair. Traditional vehicle damage assessment mainly relies on manual assessment, and the assessor needs to go to the scene in person or rely on pictures, videos and other information to make judgments. This method is not only time-consuming and labor-intensive, but also easily affected by human factors, resulting in subjectivity and inconsistency in the assessment results.

[0003] With the development of technology, vehicle damage assessment has gradually introduced technical means such as image recognition, artificial intelligence, and big data in order to improve the efficiency and accuracy of the damage assessment process. At present, vehicle damage assessment technology based on artificial intelligence has been initially applied to some insurance companies. These technologies can automatically analyze auto insurance pictures, identify the damaged parts and types of vehicles, and generate preliminary damage assessment reports. However, the existing technology still has some shortcomings in practical applications.

[0004] First, the speed of uploading, processing and analyzing images is slow, especially during peak hours. Delays in image processing will extend the entire claims process. Second, the accuracy of the damage assessment results depends on the training data of the model. When faced with new or complex damages, the generalization ability of the existing model is insufficient, which may lead to misjudgment. In addition, when processing a large number of auto insurance images, the existing technology has an imperfect allocation and scheduling mechanism for computing resources, which is prone to task accumulation and processing bottlenecks. Finally, although some systems have begun to apply cloud computing and distributed technologies, there are still hidden dangers in the security of data transmission and storage, especially when it comes to user privacy and insurance data, data security is particularly important.

[0005] Therefore, how to further optimize the efficiency and accuracy of vehicle damage assessment, how to improve the response speed of the system through distributed processing and intelligent analysis technology, and how to process massive auto insurance images while ensuring data security have become hot spots and difficulties in current technical research in the field of vehicle insurance. Summary of the invention

[0006] The main purpose of the present invention is to provide a vehicle damage assessment result query method, device, equipment and storage medium, aiming to solve the technical problems in the prior art that the vehicle damage assessment processing efficiency is low, resulting in insufficient integrity of the query results and affecting the smooth progress of the claims process.

[0007] To achieve the above object, the present invention provides a vehicle damage assessment result query method, comprising:

[0008] The vehicle claims processing end uploads the auto insurance pictures to the backend damage assessment processing end in batches through the picture request interface, and assigns a unique ID to each auto insurance picture each time it is uploaded. The ID and the upload request are recorded in the request result table, marking the receiving status of each auto insurance picture.

[0009] The background damage assessment processing end extracts the image download task from the request result table, completes the download of each car insurance image, and updates the receiving status of each car insurance image in the request result table;

[0010] After a certain car insurance picture is downloaded, the background damage assessment processing end adds the car insurance picture processing task of the car insurance picture to the task queue, waiting for the car insurance processing process to perform asynchronous processing;

[0011] The auto insurance image processing task is extracted from the task queue through the auto insurance processing process, and each auto insurance image is analyzed and processed to generate the corresponding damage morphology description;

[0012] According to the damage morphology description of each auto insurance picture generated, combined with the preset repair strategy set, a repair plan corresponding to each auto insurance picture is generated, and the repair plan is stored together with the damage result;

[0013] Based on the user's query request, the damage morphology description and corresponding repair plan of each car insurance picture are displayed to the user.

[0014] In one embodiment, the auto insurance processing process extracts the auto insurance picture processing task from the task queue, analyzes and processes each auto insurance picture, and generates a corresponding damage morphology description, including:

[0015] Extracting the pending auto insurance image processing tasks from the task queue through the auto insurance processing process, preliminarily classifying the auto insurance images, and identifying the type and structural area of ​​the vehicle, wherein the structural area includes the external structural area of ​​the vehicle body, the internal structural area of ​​the vehicle body, and the chassis area;

[0016] Based on the classification results, the various parts of the vehicle are segmented and identified to determine the specific body parts in the car insurance picture;

[0017] Perform damage detection on the identified body parts, determine the type and location of the damage, and record the damage information of the body parts;

[0018] Analyze the positional relationship between the damage and the body parts, determine the damage morphology of the damaged body parts, and generate the corresponding damage morphology description.

[0019] In one embodiment, performing damage detection on the identified vehicle body parts, determining the type and location of the damage, and recording the damage information of the vehicle body parts include:

[0020] Identify the damage types of the vehicle body parts and determine the specific location of each damage on the vehicle body parts, wherein the damage types include scratches, dents, cracks, deformations and rib line damage;

[0021] Analyze the rib damage on the body parts through the rib model to determine the length and severity of the rib damage;

[0022] Analyze the dent damage on the body parts through the dent depth model to determine the depth and impact of the dent damage;

[0023] Analyze scratches, cracks and deformation damage on body parts to determine the basic damage conditions, including length, depth, shape and severity;

[0024] Summarize all damage analysis results of body parts and record complete damage information of body parts.

[0025] In one embodiment, based on the user's query request, the damage form description and the corresponding repair solution of each car insurance picture are displayed to the user, including:

[0026] Determine the image to be queried based on the user's query request, and present the damage morphology description and the corresponding repair solution of the image to be queried to the user;

[0027] When returning the query results to the user, if it is detected that there are unqueried pictures, check whether the ID of the unqueried picture exists in the request result table of the database;

[0028] If the ID of the unqueried image does not exist in the request result table, an upload or download error code is returned, and the non-existent image to be queried is ignored in the query result finally returned to the user;

[0029] If the ID of the unqueried image exists in the request result table, check the interval between the upload request time of the unqueried image and the current query time;

[0030] If the time interval exceeds the preset time interval, the processing task of the image to be queried is put back into the task queue, and a processing error code is returned to prompt the user to try again later;

[0031] If the time interval does not exceed the preset time interval, a processing prompt code is returned to prompt the user to try again later;

[0032] If all the images to be queried can be found, the query results of each image to be queried are merged to generate the final query result and display it to the user.

[0033] In one embodiment, the auto insurance processing process extracts the auto insurance picture processing task from the task queue, and analyzes and processes each auto insurance picture, including:

[0034] Prioritize the auto insurance image processing tasks and generate a task sequence based on the urgency, complexity, and expected processing time of each auto insurance image processing task;

[0035] Based on the generated task sequence, processing resources are allocated to each auto insurance image processing task in order of priority;

[0036] Each auto insurance image processing task is decomposed into multiple subtasks, and multiple subtasks are executed simultaneously on a multi-core processor or a distributed computing environment.

[0037] In one embodiment, based on the damage morphology description of each auto insurance picture generated, combined with a preset repair strategy set, a repair solution corresponding to each auto insurance picture is generated, including:

[0038] Parse the damage morphology description of each car insurance image to extract the damage category and the corresponding damage severity;

[0039] Based on the extracted damage category and damage severity, the corresponding repair options are matched in the preset repair strategy set;

[0040] Determine the repair method, estimated repair time and repair cost based on the matched repair options;

[0041] Generate a repair plan that includes the repair method, estimated repair time and repair cost.

[0042] In one embodiment, based on the user's query request, the damage form description and the corresponding repair solution of each car insurance picture are displayed to the user, including:

[0043] Generate a corresponding virtual 3D vehicle model based on the damage morphology description of each car insurance picture;

[0044] Mark each damaged part on the virtual 3D vehicle model, and set different colors for different parts of the vehicle according to the damage category and severity in the damage morphology description to distinguish the degree of damage;

[0045] Marking a corresponding repair plan next to the damaged part, wherein the repair plan includes a repair method, an estimated repair time and a repair cost;

[0046] The virtual 3D vehicle model is displayed through the user interface, allowing users to rotate and zoom the model to view the damage at different angles and the corresponding repair plan.

[0047] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a vehicle damage assessment result query device, which includes a memory, a processor, and a vehicle damage assessment result query program stored on the memory and executable on the processor, and when the vehicle damage assessment result query program is executed by the processor, the steps of the vehicle damage assessment result query method as described above are implemented.

[0048] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer storage medium, on which a vehicle damage assessment result query program is stored, and when the vehicle damage assessment result query program is executed by a processor, the steps of the vehicle damage assessment result query method as described above are implemented.

[0049] Beneficial effects: The present invention relates to a vehicle damage assessment result query method, in which the vehicle claims processing end uploads the auto insurance pictures to the background damage assessment processing end in batches, and records the picture ID and the upload request in the request result table each time the upload is made, marking the receiving status of each auto insurance picture. The background damage assessment processing end extracts the picture download task from the request result table, completes the download of each auto insurance picture, and updates the receiving status. After the download is completed, the auto insurance picture processing task is added to the task queue, and the auto insurance picture processing task is extracted from the task queue through the auto insurance processing process to generate the corresponding damage morphology description. Based on the generated damage morphology description, the corresponding repair plan is generated in combination with the repair strategy set, and the repair plan and the damage result are stored. When the user queries, the system displays the damage morphology description and the corresponding repair plan of each auto insurance picture. The present invention improves the processing efficiency of the damage assessment process, ensures the integrity of the query results, and makes the claims process more efficient and smooth. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0051] Figure 1 A schematic diagram of a flow chart of an embodiment of a method for querying vehicle damage assessment results according to the present invention;

[0052] Figure 2 This is a functional module diagram of a preferred embodiment of the vehicle damage assessment result query device of the present invention;

[0053] Figure 3 This is a schematic diagram of the structure of the device hardware operating environment involved in the vehicle damage assessment result query device embodiment of the present invention. DETAILED DESCRIPTION

[0054] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0055] In modern society, with the continuous increase in the number of cars, vehicle insurance has become one of the important means to protect the rights and interests of car owners. After a traffic accident, the car owner usually claims through the insurance company. One of the core links in the claims process is vehicle damage assessment. Vehicle damage assessment refers to the assessment of the damage to the damaged vehicle after the traffic accident and the determination of the cost of repair. Traditional vehicle damage assessment mainly relies on manual assessment, and the assessor needs to go to the scene in person or rely on pictures, videos and other information to make judgments. This method is not only time-consuming and labor-intensive, but also easily affected by human factors, resulting in subjectivity and inconsistency in the assessment results.

[0056] With the development of technology, vehicle damage assessment has gradually introduced technical means such as image recognition, artificial intelligence, and big data in order to improve the efficiency and accuracy of the damage assessment process. At present, vehicle damage assessment technology based on artificial intelligence has been initially applied to some insurance companies. These technologies can automatically analyze auto insurance pictures, identify the damaged parts and types of vehicles, and generate preliminary damage assessment reports. However, the existing technology still has some shortcomings in practical applications.

[0057] First, the speed of uploading, processing and analyzing images is slow, especially during peak hours. Delays in image processing will extend the entire claims process. Second, the accuracy of the damage assessment results depends on the training data of the model. When faced with new or complex damages, the generalization ability of the existing model is insufficient, which may lead to misjudgment. In addition, when processing a large number of auto insurance images, the existing technology has an imperfect allocation and scheduling mechanism for computing resources, which is prone to task accumulation and processing bottlenecks. Finally, although some systems have begun to apply cloud computing and distributed technologies, there are still hidden dangers in the security of data transmission and storage, especially when it comes to user privacy and insurance data, data security is particularly important.

[0058] Therefore, how to further optimize the efficiency and accuracy of vehicle damage assessment, how to improve the response speed of the system through distributed processing and intelligent analysis technology, and how to process massive auto insurance images while ensuring data security have become hot spots and difficulties in current technical research in the field of vehicle insurance.

[0059] See also Figure 1 , Figure 1 This is a flow chart of an embodiment of a vehicle damage assessment result query method provided by the present invention. It should be noted that although a logical sequence is shown in the flow chart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0060] like Figure 1 As shown, the vehicle damage assessment result query method proposed by the present invention includes the following steps:

[0061] S10, the vehicle claims processing end uploads the auto insurance pictures to the backend damage assessment processing end in batches through the picture request interface, and assigns a unique ID to each auto insurance picture at each upload, records the ID and the upload request in the request result table, and marks the receiving status of each auto insurance picture;

[0062] In this embodiment, the vehicle claims processing end refers to a front-end system or device used by users or insurance claims personnel to collect and upload pictures related to vehicle damage. The processing end may include a smartphone, tablet computer or other device with picture collection and upload functions, which is responsible for uploading the car insurance pictures taken on site to the back-end damage assessment processing end.

[0063] The image request interface is a communication channel between the front-end system and the back-end damage assessment processing end, and is used to receive and transmit auto insurance image data. Through this interface, the front-end system uploads auto insurance images one by one to the back-end damage assessment processing end, and generates and transmits metadata related to the image (such as image ID, upload time, image type, etc.) during the upload process.

[0064] In order to improve system processing efficiency and reduce bandwidth usage, the upload of auto insurance pictures is divided into several batches. This allows for effective management of the upload process when processing resources are limited or network conditions are unstable, avoiding performance bottlenecks caused by a single large batch of data transfers.

[0065] The backend damage assessment processing end refers to the core system that processes auto insurance pictures during the vehicle claims process, and is responsible for receiving, storing, and processing uploaded auto insurance pictures. This backend system usually runs on a server or cloud computing platform and has the ability to process large-scale image data.

[0066] Some large insurance companies will have their own damage assessment processing terminals, which are combined with their own claims systems and databases for damage assessment processing. These systems usually integrate specific algorithms and data processing modules to provide highly customized functions. Many companies also use cloud computing platforms (such as AWS, Google Cloud, and Alibaba Cloud) to build backend damage assessment processing terminals, using the storage and computing power of the cloud to process a large number of auto insurance pictures. Some companies also rely on damage assessment processing platforms provided by third parties, such as APIs and backend processing capabilities provided by intelligent damage assessment service providers. These platforms usually use AI technology, image recognition, deep learning and other technologies to process auto insurance pictures. Some damage assessment processing systems may use a distributed computing architecture to distribute processing tasks to multiple server nodes to improve processing speed and reliability.

[0067] The request result table is a data storage structure in the background damage assessment processing end, which is used to record the request information of each image upload, including the image ID, upload time, image status, etc. This table ensures that all uploaded images can be processed correctly by tracking the upload and processing status of each auto insurance image.

[0068] When each auto insurance image is uploaded, the system will assign it a unique identifier (ID). This ID is used to uniquely identify each image, ensure that the processing status of the image can be accurately tracked during subsequent processing, and avoid confusion between images. The ID is usually automatically generated and can be a number, letter, or a combination thereof.

[0069] After the auto insurance image is successfully uploaded to the backend damage assessment processing end, the system will update the relevant records in the request result table to mark that the image has been successfully received. This mark helps the system track the processing progress of the image in subsequent steps and reprocess failed upload tasks when necessary.

[0070] In one specific implementation, a network status monitoring module is used to monitor the bandwidth and delay of the network in real time during the upload process. According to the monitoring results, the number of pictures uploaded each time is dynamically adjusted. If the network condition is good, the number of pictures uploaded in a single batch is increased; if the network condition is poor, the number of pictures in each batch is reduced. During the upload process, the system not only records the receiving status of each picture, but also records the network status data such as the bandwidth and delay of each batch uploaded. These data will be used for adaptive adjustment of subsequent upload processes to further optimize the upload strategy. After each picture upload is completed, the system automatically analyzes the receiving status. If the reception fails, the system adjusts the retransmission strategy according to the network condition and records it in the request result table. Through the adaptive upload strategy, the upload efficiency can be optimized under different network conditions, the upload failure rate can be reduced, and the overall upload success rate and processing efficiency can be improved.

[0071] By uploading auto insurance pictures in batches and recording the receiving status of each picture during the uploading process, the success rate and processing efficiency of picture uploading can be effectively improved, and upload failures caused by network fluctuations or system delays can be reduced. At the same time, by tracking the status of the request result table, the system can accurately monitor the uploading and receiving process of each picture, ensuring that all pictures are processed in a timely manner in the subsequent processing steps, and enhancing the documentation and integrity of the query results.

[0072] S20, the background damage assessment processing end extracts the image download task from the request result table, completes the download of each car insurance image, and updates the receiving status of each car insurance image in the request result table;

[0073] In this embodiment, the background damage assessment processing end is the core system for processing auto insurance pictures, which usually runs on a server or cloud computing platform. The background damage assessment processing end is responsible for receiving uploaded pictures and extracting the picture download tasks to be processed according to the records in the request result table. It is the processing center of the entire vehicle damage assessment process.

[0074] The request result table is a database table that records detailed information about auto insurance image upload requests, including each image's ID, upload time, and receiving status. The backend damage assessment processing end determines the image tasks that need to be downloaded by querying this table.

[0075] The image download task refers to the backend extracting the pending image tasks from the request result table, and downloading these images from the front-end or intermediate storage location to the back-end damage assessment processing end for subsequent processing. This step ensures that all uploaded images can be successfully received by the back-end system.

[0076] After downloading each auto insurance image, the backend damage assessment processing end will update the relevant records in the request result table, marking the image as "received" or "downloaded successfully" so that the subsequent processing stage can proceed smoothly. This step ensures that the system can track the processing progress of each image in real time.

[0077] In the actual implementation process, the background damage assessment processing end will check the request result table regularly or in real time, extract all the image records marked as "to be downloaded" or "uploaded successfully", and generate image download tasks. The background system downloads these images to the local computer through the network or cloud storage, and after the download is completed, it updates the receiving status of the image in the request result table to ensure that each image can be correctly processed and tracked in the processing flow.

[0078] In one specific implementation method, the background damage assessment processing end uses a timed polling mechanism to regularly check the image tasks to be downloaded in the request result table to ensure the timeliness and effectiveness of task management. The background damage assessment processing end sets a timed task to check the request result table every 10 minutes, extract all image records marked as "to be downloaded", and generate corresponding image download tasks. The background system downloads the images from the front-end device or intermediate storage server to the background damage assessment processing end through a high-speed network interface, and ensures the integrity of the data during the download process. After each image is downloaded, the system immediately updates the receiving status in the request result table and marks it as "download successful". The timed polling mechanism ensures the real-time and batch processing capabilities of the image download task, is suitable for use in large-scale auto insurance case processing, and can effectively avoid task accumulation.

[0079] By extracting the image download task from the request result table and completing the download of each auto insurance image, the background damage assessment processing end can obtain the image data uploaded by the user in a timely manner and update the receiving status of each auto insurance image. This process ensures that the processing progress of each auto insurance image after uploading is effectively tracked, avoiding the loss of images or upload failures, thereby improving the reliability and data integrity of the entire damage assessment process.

[0080] S30, after a certain car insurance picture is downloaded, the background damage assessment processing terminal adds the car insurance picture processing task of the car insurance picture to the task queue, and waits for the car insurance processing process to perform asynchronous processing;

[0081] In this embodiment, the completion of downloading a certain car insurance picture means that the picture download task corresponding to a certain car insurance picture extracted from the request result table by the background damage assessment processing end has been successfully executed, and the car insurance picture has been completely downloaded to the local storage or cloud storage of the background damage assessment processing end, and has been confirmed by the system for subsequent processing.

[0082] The auto insurance image processing task refers to the task generated for each downloaded auto insurance image. This task includes all the analysis and processing steps that need to be performed on this auto insurance image, such as classification, damage detection, and repair plan generation.

[0083] The task queue is a structure used by the background damage assessment processing end to manage and schedule pending auto insurance image tasks. The task queue distributes processing tasks to the background auto insurance processing process in sequence according to a certain priority or scheduling strategy. The task queue ensures that all auto insurance image processing tasks can be executed in an orderly manner without loss or omission due to insufficient resources.

[0084] The auto insurance processing process extracts tasks from the task queue in a non-blocking manner and processes the auto insurance pictures. Asynchronous processing means that the background system can process multiple tasks at the same time without having to wait for a task to be completed before executing the next task. This processing method improves the system's parallel processing capabilities and overall efficiency.

[0085] After the image is downloaded, the background damage assessment processing end will immediately generate a corresponding processing task and add it to the task queue. The task queue distributes the tasks to the processing process in turn according to the set scheduling strategy. The processing process extracts the tasks asynchronously and performs classification, damage detection, and repair plan generation on each car insurance image. After the processing is completed, the results are stored in the system for use in subsequent steps.

[0086] In one specific implementation, after the image is downloaded, the system generates a processing task and monitors the system's computing resource status (such as CPU, memory, etc.) in real time. According to the current resource usage, the position of the task in the queue is dynamically adjusted to ensure that the system processes more tasks when the resource load is low. When the resource load of the auto insurance processing process is low, multiple tasks are extracted in batches for parallel processing to improve the overall processing efficiency. It can optimize the system's resource utilization, avoid resource overload, and improve the system's parallel processing capabilities.

[0087] By adding the auto insurance image processing task to the task queue and waiting for asynchronous processing, the background damage assessment processing end can efficiently manage and schedule the image processing tasks, ensuring that each image can be processed as soon as possible if system resources allow. The asynchronous processing method improves the system's parallel processing capabilities, avoids task blocking and resource waste, thereby improving the overall efficiency of the damage assessment process and ensuring the timeliness and accuracy of the damage assessment results.

[0088] S40, extracting the auto insurance picture processing task from the task queue through the auto insurance processing process, analyzing and processing each auto insurance picture, and generating a corresponding damage morphology description;

[0089] In this embodiment, the auto insurance processing process is usually composed of multiple independent sub-processes or threads, which can process multiple tasks from the task queue in parallel. The process can include multiple algorithm models, such as image classification, damage detection, feature extraction, etc., for comprehensive analysis of auto insurance pictures.

[0090] The task queue is a data structure in the system that stores the auto insurance image tasks to be processed. The task queue is sorted according to a certain scheduling strategy, and the auto insurance processing process extracts tasks from the queue in sequence for processing. The management of the task queue ensures that all auto insurance image processing tasks can be executed in an orderly and efficient manner.

[0091] The auto insurance image processing task is a specific task generated for each auto insurance image. This task includes multiple steps, such as image preprocessing, region segmentation, damage detection and morphological analysis. Each task is automatically generated by the system and added to the task queue, waiting for the auto insurance processing process to process it.

[0092] Analysis and processing refers to a series of processing operations on auto insurance images, including but not limited to image preprocessing (such as denoising and enhancement), target detection (such as identifying damaged areas on the vehicle body), feature extraction and classification. These processing operations rely on modern computer vision methods such as deep learning algorithms and image processing technology.

[0093] The damage morphology description is the final output of the analysis and processing, which contains an accurate description of the damaged area in each car insurance image. The description includes the location, type (such as scratches, dents, cracks, etc.), severity and affected body parts of the damage. This information will be used to generate repair plans later.

[0094] During system operation, the auto insurance processing process extracts auto insurance image processing tasks from the task queue and analyzes and processes each image step by step. First, the system pre-processes the image to optimize image quality; then, the algorithm model performs regional segmentation and damage detection on the image to identify the specific location and type of damage; finally, the system generates a complete description of the damage morphology, including information such as damage type, location, severity, etc., and stores it for subsequent processing.

[0095] In one specific implementation, the auto insurance processing process extracts the auto insurance image processing task from the task queue. The system first performs edge detection to identify the outline and damaged area of ​​the vehicle; then accurately locates the damaged area through the region growing algorithm; and finally uses texture analysis to determine the type and severity of the damage. The results of the multiple model outputs are combined to generate a damage morphology description, including the specific damage type, location, and severity.

[0096] By extracting the auto insurance image processing tasks from the task queue and analyzing them through the auto insurance processing process, the system can efficiently and accurately analyze the damage morphology of each auto insurance image and generate a detailed damage description. This significantly improves the accuracy and efficiency of damage assessment, ensuring that each auto insurance image can be processed in a timely and accurate manner, thereby improving the reliability of subsequent repair plan generation and the integrity of query results.

[0097] S50, generating a repair plan corresponding to each car insurance picture according to the damage morphology description of each car insurance picture generated and combining with a preset repair strategy set, and storing the repair plan together with the damage result;

[0098] In this embodiment, the damage morphology description generated for each car insurance picture refers to the detailed damage morphology information generated after analyzing and processing the car insurance picture through the previous processing steps. This information includes the specific type of damage (such as scratches, dents, cracks, etc.), the location of the damage, the severity, and the damaged parts. These descriptions provide basic data for the subsequent generation of repair solutions.

[0099] A repair strategy set is a predefined set of repair solution rules based on industry standards, manufacturer recommendations or historical repair data, covering repair methods for different damage types and severities. A repair strategy set may include different repair techniques (such as sheet metal, painting, part replacement, etc.), required repair time, materials and cost estimates, etc.

[0100] According to the damage morphology description, the corresponding repair method is matched in the repair strategy set to generate a specific repair plan for each car insurance picture. The repair plan usually includes information such as repair steps, required time, materials, estimated cost, etc. to ensure that the damage is properly handled.

[0101] The generated repair plan is stored in the system database together with the previous damage morphology description, so that in subsequent query or processing steps, the system can easily access and display the complete damage assessment and repair information.

[0102] In the system, after the auto insurance processing generates a description of the damage morphology of each auto insurance image, the system automatically calls the preset repair strategy set. Based on the type, location and severity of the damage, the system selects the most suitable repair strategy and generates a specific repair plan. Finally, the repair plan is stored in the database together with the damage morphology description for subsequent query, display and repair execution.

[0103] In one specific implementation, a personalized repair strategy is introduced to generate a repair plan based on the usage history and damage record of a specific vehicle. The system parses the damage morphology description and analyzes it in combination with the vehicle's historical repair record and usage. The system selects a personalized repair strategy from the repair strategy set based on the vehicle's specific historical data. For example, for a part that has been repaired multiple times, a more thorough repair method may be selected. The system generates a personalized repair plan, taking into account the special conditions of the vehicle and formulating more precise repair steps and material usage. The personalized repair plan is stored together with the damage morphology description to provide a tailored repair plan for a specific vehicle. The personalized repair strategy set improves the pertinence and accuracy of the plan, and is particularly suitable for vehicles with special repair needs.

[0104] By combining the damage morphology description of each generated auto insurance image with the preset repair strategy set to generate the corresponding repair plan, the system can provide accurate repair suggestions for each vehicle damage and store the repair plan together with the damage results. This improves the accuracy and practicality of the repair plan, ensures that the vehicle can be properly repaired, and provides a complete record for subsequent inquiries and processing. This method greatly improves the automation level of vehicle damage assessment and repair processes, and improves user satisfaction and claims efficiency.

[0105] S60: Based on the user's query request, the damage form description and the corresponding repair plan of each car insurance picture are displayed to the user.

[0106] In this embodiment, the user's query request refers to a query operation initiated by the user (usually the car owner or insurance claims adjuster) through the system, requesting the system to return the damage assessment result and repair plan for a specific car insurance case. The query request usually contains the case number, image ID or other relevant identifiers to accurately locate the information required by the user.

[0107] The system presents the generated damage morphology description and repair plan to the user in a visual way through the user interface or other interactive means. The display content can be in the form of text, charts, images, etc., aiming to enable the user to intuitively understand the damage condition and repair plan of the vehicle.

[0108] After the user initiates a query request, the system extracts the damage morphology description and repair plan of the car insurance pictures related to the query request from the database. The system integrates this information and displays it to the user through the user interface. The display can be in the form of text description, image annotation or 3D model, so that the user can intuitively understand the damage of the vehicle and the corresponding repair measures.

[0109] In one embodiment, the system extracts all damage morphology descriptions and repair solutions related to the query request. The system generates a detailed PDF report that contains a text description of the damage, image annotations, and detailed instructions for the repair solution. The user interface provides a link to download the report, and the user can save or print the report. The PDF report provides complete and detailed damage and repair information, which is suitable for scenarios that need to be archived or shared.

[0110] By displaying the damage form description and corresponding repair plan of each car insurance picture to the user based on the user's query request, the system can allow users to intuitively and clearly understand the damage and repair plan of the vehicle. The display process helps users quickly grasp the damage assessment results and provide a reliable basis for repair decisions. It enhances the user experience, improves the visualization of damage assessment information, shortens the time of the claims process, and improves the efficiency and quality of the overall claims service.

[0111] The present invention relates to a vehicle damage assessment result query method, in which a vehicle claim processing end uploads auto insurance pictures to a background damage assessment processing end in batches, records the picture ID and the upload request in a request result table each time the picture is uploaded, and marks the receiving status of each auto insurance picture. The background damage assessment processing end extracts the picture download task from the request result table, completes the download of each auto insurance picture, and updates the receiving status. After the download is completed, the auto insurance picture processing task is added to the task queue, and the auto insurance picture processing task is extracted from the task queue through the auto insurance processing process to generate a corresponding damage morphology description. Based on the generated damage morphology description, a corresponding repair plan is generated in combination with a repair strategy set, and the repair plan and the damage result are stored. When a user queries, the system displays the damage morphology description and the corresponding repair plan of each auto insurance picture. The present invention improves the processing efficiency of the damage assessment process, ensures the integrity of the query results, and makes the claims process more efficient and smooth.

[0112] In one embodiment, the above S40 includes:

[0113] S401, extracting to-be-processed auto insurance image processing tasks from a task queue through an auto insurance processing process, preliminarily classifying the auto insurance images, and identifying the type and structural area of ​​the vehicle, wherein the structural area includes an external structural area of ​​the vehicle body, an internal structural area of ​​the vehicle body, and a chassis area;

[0114] S402, based on the classification result, segment and identify various parts of the vehicle to determine specific vehicle body parts in the vehicle insurance picture;

[0115] S403, performing damage detection on the identified body parts, determining the type and location of the damage, and recording the damage information of the body parts;

[0116] S404, analyzing the positional relationship between the damage and the body component, determining the damage form of the damaged body component, and generating a corresponding damage form description.

[0117] In this embodiment, the task queue is a structure sorted in chronological order or priority order, storing all pending auto insurance picture tasks. The task queue can be implemented as a first-in-first-out (FIFO) queue, or it can be managed based on priority (Priority Queue). When the system calls the auto insurance processing process, the auto insurance processing process extracts the pending auto insurance picture tasks from the head of the task queue or in priority order. Task extraction usually uses a thread or process scheduling mechanism to ensure that the processing process can extract tasks in parallel or sequentially in a multi-tasking environment to avoid task loss or repeated processing.

[0118] Before classification, the system preprocesses the image, such as denoising, image enhancement, and grayscale processing. These operations can improve the accuracy of subsequent classification. The system uses image classification algorithms (such as convolutional neural networks, CNN) to preliminarily classify auto insurance images. The classification results include the type of vehicle (such as sedans, SUVs, trucks, etc.) and structural areas (such as front body, rear body, etc.). This classification process is usually performed using a pre-trained model, and the classification model can be trained based on a large amount of labeled data to improve recognition accuracy. The system further identifies the main structural areas of the vehicle. Structural areas refer to parts of the vehicle with specific functions or physical separations. These parts are used for detailed analysis and damage location during the vehicle damage assessment process. Structural areas can be identified by segmentation of function, physical structure, or vehicle design, including:

[0119] External structural area of ​​the vehicle body: front (such as hood, front bumper, headlights, grille); side (such as doors, side skirts, side mirrors, windows); rear (such as trunk lid, rear bumper, taillights).

[0120] Internal structural area of ​​the vehicle body: cockpit area (seats, steering wheel, instrument panel, airbags); engine compartment area (engine, radiator, transmission system).

[0121] Chassis: Chassis (frame, floor, suspension system).

[0122] In the damage assessment process, the structural area is used to help the system determine the specific damage location and type based on the car insurance picture, so as to further identify the damaged parts. Each structural area represents a functional partition or physical block of the vehicle. When analyzing, the system will adopt different damage detection methods according to these areas to ensure the accuracy of the damage assessment results.

[0123] The system uses image segmentation algorithms (such as semantic segmentation, edge detection, etc.) to segment the image into components and separate the various components of the vehicle (such as doors, hoods, windows, etc.) from the overall image. Segmentation algorithms usually combine classification results and apply different segmentation strategies to specific structural areas. Based on the segmentation, the system identifies each component and determines the specific component type and location. This step is usually performed using pattern recognition technology or rule-based recognition methods. The recognition results include the name of the component (such as the left front door, right rear window, etc.) and its position coordinates in the image. Through segmentation and recognition, complex vehicle images are decomposed into manageable components, which provides a more refined operational target for subsequent damage detection.

[0124] The system uses specialized damage detection algorithms (such as feature matching-based detection, deep learning detection, etc.) to perform damage detection on the identified body parts. Damage detection includes common types of vehicle damage such as scratches, dents, cracks, and deformations. The detection algorithm can identify the edges, shapes, colors, and other features of these damages. The detection algorithm not only identifies the location of the damage, but also determines the specific type of damage. The system associates the detection results with specific damage types by matching predefined damage type models. The system records the location coordinates of the damage and associates it with the information of the body parts. This data is usually stored in a database in a structured form for subsequent analysis and processing. All detected damage information, including type, location, severity, etc., will be stored in the system. This information will be used to generate a description of the damage morphology.

[0125] The system analyzes the spatial relationship between the damage and the body parts to determine the scope of the damage. For example, whether the damage involves the key structure of the body, whether it will affect the overall performance of the vehicle, etc. Positional relationship analysis can use spatial analysis algorithms, combined with the three-dimensional model of the vehicle, to accurately evaluate the impact of the damage on the body. Based on the positional relationship analysis, the system determines the overall damage morphology of the damaged parts. The damage morphology description includes the geometry of the damage, the scope of impact, the potential impact on vehicle performance, etc. The system generates a complete damage morphology description and formats it as structured data or natural language description. These descriptions will be used for subsequent repair plan formulation and user presentation. By analyzing the positional relationship between the damage and the body parts, the system is able to generate a comprehensive damage morphology description to ensure that the vehicle's damage information is accurate and detailed, providing a basis for repair decisions.

[0126] This embodiment extracts the auto insurance image processing task from the task queue through the auto insurance processing process, analyzes and processes each auto insurance image, and the system can accurately generate a description of the vehicle's damage morphology. This process includes steps such as classification, segmentation, recognition, and damage analysis to ensure that the damage information of each auto insurance image is fully and accurately described. The resulting damage morphology description provides a reliable basis for the formulation and presentation of subsequent repair plans, improving the accuracy and efficiency of vehicle damage assessment.

[0127] In one embodiment, the above S403 includes:

[0128] S4031, identifying the damage type of the body component and determining the specific location of each damage on the body component, wherein the damage type includes scratches, dents, cracks, deformation and rib line damage;

[0129] S4032, analyzing the rib damage on the body parts through the rib model to determine the length and severity of the rib damage;

[0130] S4033, analyzing the dent damage on the vehicle body parts by using a dent depth model to determine the depth and impact of the dent damage;

[0131] S4034, analyze scratches, cracks and deformation damage on body parts to determine the basic damage conditions, including length, depth, morphology and severity;

[0132] S4035, summarizing all damage analysis results of the body parts and recording complete damage information of the body parts.

[0133] In this embodiment, the system uses image processing and pattern recognition algorithms to detect the type of damage on the body parts. Common types of damage include scratches, dents, cracks, deformations, and rib damage. Recognition algorithms usually rely on pre-trained models that can determine the specific type of damage by analyzing features such as texture, shape, color, etc. in the image. After identifying the type of damage, the system further determines the specific location of each type of damage on the body part. The position data is usually expressed in the form of coordinates, identifying the specific area of ​​the damage in the body part. Position determination can be achieved through algorithms such as edge detection and region growing to ensure accurate positioning of the damaged area.

[0134] The rib model is a model specifically designed for analyzing ribs in a vehicle body. It can identify and analyze rib damage on a vehicle body component, such as rib breakage, distortion, or indentation. The rib model usually combines edge detection and shape analysis algorithms to assess the integrity of the rib. The system measures the length of the rib damage to determine the physical size of the damage. The severity is assessed based on the degree of breakage or deformation of the rib. The length measurement can be calculated by pixel distance, while the severity is usually assessed by analyzing the impact of the damage on the overall structure of the rib.

[0135] The dent depth model is a model specifically used to evaluate dent damage on a vehicle body. It determines the depth of the dent by analyzing the morphological features and light and shadow changes of the image. The model is usually based on 3D reconstruction technology or morphological analysis technology, and accurately measures the depth of the dent by analyzing the shape and light and shadow relationship of the dent area. Depth measurement determines the physical depth of the dent by performing 3D reconstruction or morphological analysis on the image data of the dent area. The impact assessment determines the overall impact of the dent on the vehicle body components based on the depth, area and functional importance of the component where the dent is located.

[0136] The system detects scratches and cracks through image analysis algorithms to identify the morphological characteristics of the damage. Morphological characteristics include the length, width and depth of the damage. The analysis of scratches usually relies on edge detection algorithms, while the analysis of cracks may be combined with crack detection technology. The analysis of deformation damage is mainly completed through shape recognition algorithms. The system identifies the deformed area of ​​the body component and measures the degree of deformation (such as bending angle, area change, etc.). The system also evaluates the impact of deformation damage on the component structure and determines the severity of the damage. The system summarizes all analysis results of scratches, cracks and deformations to form a description of the basic damage situation. The description includes length, depth, morphology and severity.

[0137] The system summarizes all the above analysis results and generates a complete body component damage report. The report contains all the identified damage types and their detailed descriptions. The summary process also ensures the accuracy and consistency of the damage information to avoid duplicate records or omissions of important information. The system records the complete damage information in the database as the basic data for subsequent repair decisions and user presentation. The damage information record is usually in a structured data format to facilitate subsequent query and processing.

[0138] By performing comprehensive damage detection on the identified body parts, this embodiment of the system can accurately identify and record various damage types and locations. The refined analysis process (such as the application of rib line model and dent depth model) ensures the accuracy and completeness of the damage information. The final summary of the damage information provides a solid foundation for the formulation of subsequent repair plans and improves the accuracy and efficiency of the vehicle damage assessment process. This processing method can improve the automation level of vehicle damage assessment, reduce human errors, and ensure the reliability of damage assessment results.

[0139] In one embodiment, the above S60 includes:

[0140] S601, determining a to-be-queried image based on a query request of a user, and presenting a description of the damage morphology of the to-be-queried image and a corresponding repair solution to the user;

[0141] S602, when returning the query result to the user, if it is detected that there is an unqueried picture, check whether the ID of the unqueried picture exists in the request result table of the database;

[0142] S603: If the ID of the unqueried image does not exist in the request result table, an upload or download error code is returned, and the non-existent image to be queried is ignored in the query result finally returned to the user;

[0143] S604, if the ID of the unqueried image exists in the request result table, check the interval between the upload request time of the unqueried image and the current query time;

[0144] S605, if the time interval exceeds the preset time interval, the processing task of the image to be queried is put back into the task queue, and a processing error code is returned to prompt the user to try again later;

[0145] S606, if the time interval does not exceed the preset time interval, return a processing prompt code to prompt the user to try again later;

[0146] S607: If all the images to be queried can be found, the query results of each image to be queried are merged to generate a final query result and displayed to the user.

[0147] In this embodiment, the system receives a query request from a user, which usually includes identification information such as a case number or a picture ID, which is used to determine the auto insurance picture that the user wants to query. The system parses the query request, extracts the identifier of the picture to be queried, and searches it in the database. The system retrieves the corresponding damage morphology description and repair plan from the database based on the identifier in the query request. The system displays the retrieved damage morphology description and repair plan to the user through the user interface, usually in the form of text, images or charts.

[0148] When the system generates query results, it checks whether there are any images that the user requested but failed to find. The system identifies the unqueried images by comparing the image IDs in the user's query request with the image ID list in the actual query results. For the detected unqueried images, the system further checks whether the IDs of these images exist in the request result table of the database. The request result table usually records the IDs of all images that have been successfully uploaded and entered the processing flow. The system determines the upload or download status of the image by querying this table.

[0149] If the system finds that the ID of the unqueried image does not exist in the request result table, it indicates that an error may have occurred during the upload or download process of the image. The system generates and returns a specific error code (such as upload failure or download failure) to prompt the user that the image may have failed at an early stage. When generating the final query results, the system ignores these non-existent image IDs to avoid misleading the user with error information. The query results that the user finally obtains only include the successfully queried images and their corresponding damage morphology descriptions and repair solutions.

[0150] For unqueried images in the request result table, the system further checks the time interval between the timestamp of the image request and the current query time. The timestamp is usually recorded when the image is uploaded and is used to identify the time when the image enters the processing flow. The system determines whether the image may not be queried in time due to long processing time based on the preset time interval threshold. This time interval check can help the system identify processing anomalies or delays. The time interval check ensures that the system can identify those images that were not queried in time due to processing delays, providing a basis for subsequent processing.

[0151] If the system finds that the time interval exceeds the preset threshold, indicating that an exception or delay may have occurred in the image processing, the system will put the processing tasks of these images back into the task queue. The operation of putting them back into the task queue ensures that the images can be processed again to avoid task failures caused by temporary errors. The system generates and returns a processing error code, prompting the user to try the query operation again later. The error code is usually designed to include detailed information to indicate the possible cause of the user's problem and the recommended follow-up actions. Through rescheduling and error code return, the system can effectively deal with delays or exceptions in image processing and ensure that images can be processed and queried in a timely manner.

[0152] If the time interval is within the normal range, the system will confirm that the image is still being processed and has not yet completed the processing task. The system will not reschedule the task of the image, but will continue to monitor its processing status. The system returns a processing prompt code to inform the user that the current image is being processed and recommends that the user try the query operation again later. This prompt code usually comes with a recommended re-query time or waiting time.

[0153] When the system confirms that all the images to be queried have been processed and successfully queried, the system will merge the damage morphology descriptions and repair solutions of each image. The fusion process may include data integration, sorting, or generating a comprehensive report to ensure that users can view the processing results of all images at a glance. The system displays the fused query results to the user through the user interface. The display format may include text reports, image annotations, charts, or interactive 3D models. Users can view the damage morphology description of each image and its corresponding repair solution in the final results to obtain comprehensive damage assessment information. Through the fusion and display of query results, the system can provide users with complete and clear information on vehicle damage and repair solutions, improving user experience and query efficiency.

[0154] Through the above steps, this embodiment can efficiently process the user's query request and accurately display the damage morphology description and repair plan of the car insurance picture. By detecting and processing the unqueried pictures, the system improves the reliability of the query results, and through the time interval check and error handling mechanism, ensures that each picture can be processed in a timely and accurate manner. Finally, the system provides users with comprehensive query results, improving the user experience and satisfaction in the vehicle claims process.

[0155] In one embodiment, the above S40 includes:

[0156] S405, extracting the auto insurance picture processing task from the task queue through the auto insurance processing process, and analyzing and processing each auto insurance picture, including:

[0157] S406, prioritizing the auto insurance image processing tasks according to the urgency, complexity, and expected processing time of each auto insurance image processing task, and generating a task sequence;

[0158] S407, allocating processing resources to each auto insurance image processing task in order of priority based on the generated task order;

[0159] S408, decomposing each auto insurance image processing task into multiple subtasks, and executing the multiple subtasks simultaneously in a multi-core processor or a distributed computing environment.

[0160] In this embodiment, the system first evaluates the attributes of each auto insurance image processing task, including the urgency, complexity, and expected processing time of the task. It may be based on external factors (such as customer demand, insurance company processing time limit) or internal policies (such as the priority of other tasks in the processing queue). The system determines the complexity level of the task by analyzing the characteristics of the auto insurance image (such as the resolution of the image, the complexity of the damage, etc.). Based on historical data or model estimates, the processing time that each task may require. Based on the above evaluation results, the system prioritizes the tasks according to predetermined weights or rules. Sorting can be combined with multiple strategies, for example: processing urgent and low-complexity tasks first. The sorting results generate an ordered task list, and the tasks are arranged from high to low priority to ensure that the most important or urgent tasks are processed first. Priority sorting ensures that the system can reasonably allocate resources according to the actual needs and characteristics of the task, optimize the order of task execution, and improve overall processing efficiency.

[0161] The system allocates processing resources to each task in turn according to the priority sorting results. Computing resources include hardware resources such as CPU, memory, GPU, and related software processing resources. During the resource allocation process, the system dynamically adjusts the resource allocation strategy to adapt to changes in tasks. For example, high-priority tasks may obtain more computing resources. The system uses resource scheduling algorithms (such as resource pooling or dynamic load balancing) to ensure that each task receives appropriate resource support during its execution. Resource allocation can be static pre-allocation or real-time dynamic allocation, depending on the real-time performance requirements of the system and the complexity of the task. Through reasonable resource allocation, the system ensures that each task can obtain the required computing resources when executing, thereby maximizing processing efficiency and reducing task waiting time.

[0162] The system decomposes each auto insurance image processing task into multiple subtasks. Subtasks can be specific steps such as image preprocessing, damage detection, and component identification. Task decomposition is usually based on a modular processing framework to ensure that each subtask can be executed independently and in parallel with other subtasks. The system distributes the decomposed subtasks to multiple nodes in a multi-core processor or distributed computing environment for parallel execution. Each subtask runs on an independent computing core or distributed node. On a single device, multiple subtasks can be executed simultaneously on different CPU cores or GPUs to speed up the processing of tasks. In a distributed environment, subtasks are distributed to multiple computing nodes to improve processing efficiency by leveraging the computing power of the cluster. After all subtasks are executed, the system will integrate the results of each subtask to form a complete processing result. The integration process includes operations such as data aggregation and result merging to ensure that the final task processing results are complete and consistent. Through task decomposition and parallel execution, the system can significantly improve the processing efficiency of tasks, reduce processing time, and adapt to the needs of high-load and large-scale image processing.

[0163] This embodiment can achieve efficient task management and processing by prioritizing the auto insurance image processing tasks, allocating computing resources, and decomposing and executing tasks in parallel. Prioritization ensures that the most urgent and important tasks are processed first, resource allocation optimizes the efficiency of computing resources, and task decomposition and parallel execution significantly speed up task processing. Ultimately, this processing method improves the overall performance of the auto insurance loss assessment system, reduces task processing time, and meets the needs of large-scale parallel processing.

[0164] In one embodiment, the above S50 includes:

[0165] S501, based on the damage morphology description of each vehicle insurance picture generated, combined with a preset repair strategy set, a repair solution corresponding to each vehicle insurance picture is generated, including:

[0166] S502, parsing the damage morphology description of each car insurance picture to extract the damage category and the corresponding damage severity;

[0167] S503, matching corresponding repair options in a preset repair strategy set based on the extracted damage category and damage severity;

[0168] S504, determining a repair method, an estimated repair time, and a repair cost based on the matched repair options;

[0169] S505: Generate a repair plan including a repair method, an estimated repair time, and a repair cost.

[0170] In this embodiment, the system first analyzes the damage morphology description generated by each car insurance picture. The analysis process includes identifying key features in the damage description, such as damage type, location, size, etc.

[0171] Damage category extraction: The system extracts the damage category (such as scratches, dents, cracks, deformations, etc.) from the damage morphology description. This step usually uses natural language processing (NLP) or image analysis technology to automate the parsing process. The system further evaluates the severity of the damage. The severity may be based on the physical characteristics of the damage (such as length, depth, area) and its impact on the function or appearance of the vehicle. The system may combine multiple factors, such as the physical characteristics of the damage, the criticality of the damage location (such as the main support parts of the vehicle body), etc., to comprehensively evaluate the severity of the damage. By parsing and extracting the damage category and severity, the system can provide accurate input data for the formulation of subsequent repair plans.

[0172] The repair strategy set is a database of predefined repair solutions that contains repair options for different damage categories and severities. Each repair option describes in detail the repair method suitable for a specific damage type and severity. The repair strategy set is organized and indexed by multiple dimensions such as damage category, severity, and vehicle type, so that the system can quickly match suitable repair options. The system searches for matching repair options in the repair strategy set based on the damage category and severity extracted from the damage morphology description. The system can use rule-based matching algorithms or machine learning models to automatically select the most appropriate repair option to ensure that the selected repair solution is both efficient and economical. The matching process of the repair strategy set ensures that each type of damage can get the best repair solution, which helps to improve the quality and efficiency of the repair.

[0173] Based on the matched repair options, the system determines the specific repair method. For example, for minor scratches, the repair method may be polishing and waxing, while for severe dents, sheet metal repair and repainting may be required. The system can further optimize the repair method based on the vehicle's usage scenario, the available technology and equipment of the repair shop, etc. The system estimates the time and cost required for the repair based on the complexity of the repair method and the severity of the damage. Time estimates may be calculated based on historical data or specific models to ensure the accuracy of the estimate. Cost estimates are also based on factors such as repair methods, material costs, and labor costs. The system automatically generates repair cost estimates. By determining the repair method and estimating the repair time and cost, the system provides users with a detailed repair plan to help them understand the repair process and expected expenses.

[0174] The system integrates all matched repair options, determined repair methods, estimated repair time and repair cost into a complete repair plan. The generated repair plan may be stored in the form of structured data for easy display or further processing. The content includes detailed repair methods for each damage, required time, cost and possible repair steps. The repair plan generated by the system can be displayed to the user, usually through a user interface, and may be accompanied by detailed instructions or step-by-step instructions for the repair plan. The system also stores the repair plan in a database for subsequent query, modification or verification. By generating a complete repair plan, the system provides users with a transparent and detailed repair plan to help users better understand the repair process and make corresponding decisions.

[0175] This embodiment analyzes the damage morphology description of the car insurance picture and combines it with the preset repair strategy set, so that the system can generate an accurate repair plan for each damage. The repair plan includes the repair method, estimated time and cost, and provides users with comprehensive repair guidance to improve the scientificity and feasibility of the repair plan, making the vehicle repair process more efficient and transparent. Users can clearly understand each repair step and the logic behind it, which improves user experience and satisfaction.

[0176] In one embodiment, the above S60 includes:

[0177] S608, generating a corresponding virtual three-dimensional vehicle model based on the damage morphology description of each vehicle insurance picture;

[0178] S609, marking each damaged part on the virtual three-dimensional vehicle model, and setting different colors for different parts of the vehicle according to the damage category and severity in the damage morphology description to distinguish the degree of damage;

[0179] S610, marking a corresponding repair plan next to the damaged part, wherein the repair plan includes a repair method, an estimated repair time and a repair cost;

[0180] S611, displaying a virtual three-dimensional vehicle model through a user interface, allowing the user to rotate and zoom the model to view damage conditions at different angles and corresponding repair solutions.

[0181] In this embodiment, the system generates a three-dimensional model of the vehicle based on the damage morphology description of the car insurance picture. The three-dimensional model is usually generated based on a preset vehicle model library by combining the damage morphology with a standard vehicle model. The system selects a three-dimensional model that matches the actual vehicle from the preset vehicle model library, and these models may be classified by model, year, structure, etc. Based on the location and type information in the damage morphology description, the system constructs the damage part on the standard three-dimensional model and generates a virtual three-dimensional vehicle model containing damage information. The generation of the three-dimensional model provides a basis for the subsequent damage visualization and repair plan display, allowing users to intuitively view the damage to the vehicle.

[0182] The system marks each damage location on the generated 3D vehicle model. The marking process involves mapping the location information in the damage morphology description onto the 3D model so that each damage can be accurately displayed in the model. The system uses precise coordinate mapping technology to ensure that the location of the damage annotation is consistent with the actual damage location to avoid errors. In order to distinguish between damages of different types and severity, the system sets color codes for different parts of the vehicle according to the damage category and severity. For example, light colors may be used for minor scratches, while dark colors or red may be used for severe dents or cracks. The color mapping rules are preset in the system, and the choice of color is directly related to the type and severity of the damage. Users can intuitively judge the severity of the damage by color. Through color coding and damage annotation, the system provides users with an intuitive way to identify and understand the damage to the vehicle, improving the visualization of damage.

[0183] The system marks the corresponding repair solutions next to the damaged parts of the virtual 3D vehicle model. These solutions are based on the previously generated repair strategies, and the specific contents include the repair method, estimated repair time and repair cost. The repair solution annotation can be displayed in the form of a text label or a pop-up window, and the user can see the corresponding repair solution at the same time when viewing the damaged part. The system associates the repair solution of each damaged part with its corresponding 3D model annotation to ensure that the repair solution can be accurately displayed next to the corresponding damaged part. The user can click or hover on the damaged part, and the system will dynamically display detailed repair solution information to enhance the user's interactive experience. The repair solution annotation provides repair suggestions directly related to the damage information. The user can view the correspondence between the damage and the repair solution at a glance to help them make maintenance decisions.

[0184] The system displays the generated virtual 3D vehicle model through the user interface. The interface design usually includes operation functions such as rotation, zooming, and translation. Users can freely adjust the viewing angle of the model to view different parts of the vehicle. The interactive controls provided by the user interface include mouse dragging, gesture operation, slider adjustment, etc. Users can rotate or zoom the 3D model through these controls. When the user rotates or zooms the 3D model, the system dynamically updates the display of the damaged part and the repair plan, so that the user can always clearly view all information from the current perspective. The system uses real-time rendering technology to ensure that the display of the 3D model and additional information can respond to user operations smoothly and quickly. The interactive design of the user interface makes the display of the 3D model more intuitive and flexible. Users can view the vehicle damage and its corresponding repair plan in detail from different angles, enhancing the visualization of information and user experience.

[0185] Through the above steps, the system of this embodiment can generate and display a virtual three-dimensional vehicle model including damage annotations and repair solutions according to the user's query request. The visualization of damage information, color coding, and direct association display of repair solutions not only improve the accuracy of vehicle damage assessment, but also enhance the user's interactive experience. Users can rotate and scale the three-dimensional model to view the vehicle's damage and repair solutions in detail from multiple angles, making it easier to make wise repair decisions. Ultimately, the system improves users' understanding and satisfaction with vehicle damage during the auto insurance claims process through this intuitive display method.

[0186] The present invention also provides a vehicle damage assessment result query device, referring to Figure 2 , Figure 2 This is a functional module diagram of a preferred embodiment of a vehicle damage assessment result query device of the present invention. The vehicle damage assessment result query device comprises:

[0187] The picture uploading module 10 is used for the vehicle claim processing end to upload the car insurance pictures to the backend damage assessment processing end in batches through the picture request interface, and assign a unique ID to each car insurance picture at each upload, record the ID and the upload request in the request result table, and mark the receiving status of each car insurance picture;

[0188] The picture download module 20 is used for the background damage assessment processing end to extract the picture download task from the request result table, complete the download of each car insurance picture, and update the receiving status of each car insurance picture in the request result table;

[0189] The task scheduling module 30 is used for, after a certain car insurance picture is downloaded, the background damage assessment processing end adds the car insurance picture processing task of the car insurance picture to the task queue, waiting for the car insurance processing process to perform asynchronous processing;

[0190] The vehicle insurance processing module 40 is used to extract the vehicle insurance picture processing task from the task queue through the vehicle insurance processing process, analyze and process each vehicle insurance picture, and generate a corresponding damage morphology description;

[0191] A repair solution generation module 50 is used to generate a repair solution corresponding to each car insurance picture according to the damage morphology description of each car insurance picture generated, combined with a preset repair strategy set, and store the repair solution together with the damage result;

[0192] The result query module 60 is used to display the damage form description and the corresponding repair plan of each car insurance picture to the user based on the user's query request.

[0193] The specific implementation of the vehicle damage assessment result query device of the present invention is basically the same as the various embodiments of the above-mentioned vehicle damage assessment result query method, and will not be repeated here.

[0194] The present invention also provides a vehicle damage assessment result query device, such as Figure 3 As shown, the vehicle damage assessment result query device may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or it may be a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0195] Those skilled in the art will understand that Figure 3 The hardware structure of the vehicle damage assessment result query device shown in does not constitute a limitation of the vehicle damage assessment result query device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0196] like Figure 3 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a vehicle damage assessment result query program. The operating system is a program for managing and controlling the vehicle damage assessment result query device and software resources, and supports the operation of the network communication module, the user interface module, the vehicle damage assessment result query program, and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0197] exist Figure 3 In the hardware structure of the vehicle damage assessment result query device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client and communicate data with the client; the processor 1001 can call the vehicle damage assessment result query program stored in the memory 1005, and perform the same operation as the vehicle damage assessment result query method.

[0198] The specific implementation of the vehicle damage assessment result query device of the present invention is basically the same as the various embodiments of the above-mentioned vehicle damage assessment result query method, and will not be repeated here.

[0199] In addition, an embodiment of the present invention further proposes a computer storage medium, on which a vehicle damage assessment result query program is stored. When the vehicle damage assessment result query program is executed by a processor, the steps of the vehicle damage assessment result query method as described above are implemented.

[0200] The specific implementation of the computer storage medium of the present invention is basically the same as the various embodiments of the above-mentioned vehicle damage assessment result query method, and will not be repeated here.

[0201] It should be noted that if software tools or components other than those of our company appear in the embodiments of the present application, they are only used for illustrative purposes and do not represent actual use.

Claims

1. A vehicle damage assessment result query method, characterized in that: The following steps are involved: The vehicle claims processing end uploads the auto insurance pictures to the backend damage assessment processing end in batches through the picture request interface, and assigns a unique ID to each auto insurance picture each time it is uploaded. The ID and the upload request are recorded in the request result table, marking the receiving status of each auto insurance picture. The background damage assessment processing end extracts the image download task from the request result table, completes the download of each car insurance image, and updates the receiving status of each car insurance image in the request result table; After a certain car insurance picture is downloaded, the background damage assessment processing end adds the car insurance picture processing task of the car insurance picture to the task queue, waiting for the car insurance processing process to perform asynchronous processing; The auto insurance image processing task is extracted from the task queue through the auto insurance processing process, and each auto insurance image is analyzed and processed to generate the corresponding damage morphology description; According to the damage morphology description of each auto insurance picture generated, combined with the preset repair strategy set, a repair plan corresponding to each auto insurance picture is generated, and the repair plan is stored together with the damage result; Based on the user's query request, the damage morphology description and corresponding repair plan of each car insurance picture are displayed to the user.

2. The vehicle damage assessment result query method according to claim 1, characterized in that: The auto insurance image processing task is extracted from the task queue through the auto insurance processing process, and each auto insurance image is analyzed and processed to generate the corresponding damage morphology description, including: Extracting the pending auto insurance image processing tasks from the task queue through the auto insurance processing process, preliminarily classifying the auto insurance images, and identifying the type and structural area of ​​the vehicle, wherein the structural area includes the external structural area of ​​the vehicle body, the internal structural area of ​​the vehicle body, and the chassis area; Based on the classification results, the various parts of the vehicle are segmented and identified to determine the specific body parts in the car insurance picture; Perform damage detection on the identified body parts, determine the type and location of the damage, and record the damage information of the body parts; Analyze the positional relationship between the damage and the body parts, determine the damage morphology of the damaged body parts, and generate the corresponding damage morphology description.

3. The vehicle damage assessment result query method according to claim 2, characterized in that: Perform damage detection on the identified body parts, determine the type and location of the damage, and record the damage information of the body parts, including: Identify the damage types of the vehicle body parts and determine the specific location of each damage on the vehicle body parts, wherein the damage types include scratches, dents, cracks, deformations and rib line damage; Analyze the rib damage on the body parts through the rib model to determine the length and severity of the rib damage; Analyze the dent damage on the body parts through the dent depth model to determine the depth and impact of the dent damage; Analyze scratches, cracks and deformation damage on body parts to determine the basic damage conditions, including length, depth, shape and severity; Summarize all damage analysis results of body parts and record complete damage information of body parts.

4. The vehicle damage assessment result query method according to claim 1, characterized in that: Based on the user's query request, the damage form description and corresponding repair plan of each car insurance picture are displayed to the user, including: Determine the image to be queried based on the user's query request, and present the damage morphology description and the corresponding repair solution of the image to be queried to the user; When returning the query results to the user, if it is detected that there are unqueried pictures, check whether the ID of the unqueried picture exists in the request result table of the database; If the ID of the unqueried image does not exist in the request result table, an upload or download error code is returned, and the non-existent image to be queried is ignored in the query result finally returned to the user; If the ID of the unqueried image exists in the request result table, check the interval between the upload request time of the unqueried image and the current query time; If the time interval exceeds the preset time interval, the processing task of the image to be queried is put back into the task queue, and a processing error code is returned to prompt the user to try again later; If the time interval does not exceed the preset time interval, a processing prompt code is returned to prompt the user to try again later; If all the images to be queried can be found, the query results of each image to be queried are merged to generate the final query result and display it to the user.

5. The vehicle damage assessment result query method as claimed in claim 1, characterized in that: The auto insurance image processing task is extracted from the task queue through the auto insurance processing process, and each auto insurance image is analyzed and processed, including: Prioritize the auto insurance image processing tasks and generate a task sequence based on the urgency, complexity, and expected processing time of each auto insurance image processing task; Based on the generated task sequence, processing resources are allocated to each auto insurance image processing task in order of priority; Each auto insurance image processing task is decomposed into multiple subtasks, and multiple subtasks are executed simultaneously on a multi-core processor or a distributed computing environment.

6. The vehicle damage assessment result query method according to claim 1, characterized in that: Based on the damage morphology description of each auto insurance picture generated, combined with the preset repair strategy set, a repair solution corresponding to each auto insurance picture is generated, including: Parse the damage morphology description of each car insurance image to extract the damage category and the corresponding damage severity; Based on the extracted damage category and damage severity, the corresponding repair options are matched in the preset repair strategy set; Determine the repair method, estimated repair time and repair cost based on the matched repair options; Generate a repair plan that includes the repair method, estimated repair time and repair cost.

7. The vehicle damage assessment result query method according to claim 1, characterized in that: Based on the user's query request, the damage form description and corresponding repair plan of each car insurance picture are displayed to the user, including: Generate a corresponding virtual 3D vehicle model based on the damage morphology description of each car insurance picture; Mark each damaged part on the virtual 3D vehicle model, and set different colors for different parts of the vehicle according to the damage category and severity in the damage morphology description to distinguish the degree of damage; Marking a corresponding repair plan next to the damaged part, wherein the repair plan includes a repair method, an estimated repair time and a repair cost; The virtual 3D vehicle model is displayed through the user interface, allowing users to rotate and zoom the model to view the damage at different angles and the corresponding repair plan.

8. A vehicle damage assessment result query device, characterized in that: The vehicle damage assessment result query device comprises: The picture upload module is used by the vehicle claims processing end to upload the car insurance pictures in batches to the backend damage assessment processing end through the picture request interface, and assign a unique ID to each car insurance picture at each upload, record the ID and the upload request in the request result table, and mark the receiving status of each car insurance picture; The picture download module is used by the background damage assessment processing end to extract the picture download task from the request result table, complete the download of each car insurance picture, and update the receiving status of each car insurance picture in the request result table; The task scheduling module is used to add the car insurance picture processing task of the car insurance picture to the task queue after the download of a car insurance picture is completed, and wait for the car insurance processing process to perform asynchronous processing; The car insurance processing module is used to extract the car insurance image processing tasks from the task queue through the car insurance processing process, analyze and process each car insurance image, and generate the corresponding damage morphology description; A repair plan generation module is used to generate a repair plan corresponding to each car insurance picture according to the damage morphology description of each car insurance picture generated, combined with a preset repair strategy set, and store the repair plan together with the damage result; The result query module is used to display the damage morphology description and corresponding repair plan of each car insurance picture to the user based on the user's query request.

9. A vehicle damage assessment result query device, characterized in that: The vehicle damage assessment result query device includes a memory, a processor, and a vehicle damage assessment result query program stored in the memory and executable on the processor. When the vehicle damage assessment result query program is executed by the processor, the steps of the vehicle damage assessment result query method as described in any one of claims 1 to 7 are implemented.

10. A computer storage medium, characterized in that: The storage medium stores a vehicle damage assessment result query program, and when the vehicle damage assessment result query program is executed by the processor, the steps of the vehicle damage assessment result query method as described in any one of claims 1-7 are implemented.

Images