CN-122023026-A - Vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses

CN122023026ACN 122023026 ACN122023026 ACN 122023026ACN-122023026-A

Abstract

The invention relates to the technical field of AI (advanced technology), in particular to a vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses, wherein the system comprises the vehicle-mounted electronic equipment, a cloud platform and the AI glasses, and the cloud platform is in communication connection with the vehicle-mounted electronic equipment and the AI glasses and is used for data fusion, damage assessment and damage assessment decision; AI glasses are worn by the surveyor for on-site visual acquisition and interactive display; compared with the prior art that a manual on-site investigation scheme is adopted, visual inspection and photo recording are carried out by relying on a surveyor to reach an accident site, and the method has the defects of low efficiency, long processing flow and extremely easiness in delaying by environmental factors.

Inventors

HE XIAOCHEN
FANG XIANYANG

Assignees

深圳市小马数智科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260107

Claims (10)

1. The vehicle insurance damage assessment system based on the vehicle-mounted electronic equipment and the AI glasses is characterized by comprising the vehicle-mounted electronic equipment, a cloud platform and the AI glasses, and specifically comprises the following components: the vehicle-mounted electronic equipment is arranged in the vehicle and used for collecting dynamic data related to accidents; The cloud platform is in communication connection with the vehicle-mounted electronic equipment and the AI glasses and is used for data fusion, damage assessment and damage assessment decision; AI glasses are worn by surveyors for on-site vision acquisition and interactive display.
2. The vehicle insurance loss assessment system based on the vehicle-mounted electronic device and the AI glasses according to claim 1, wherein the vehicle-mounted electronic device comprises: The high-precision GNSS positioning module is used for acquiring accurate position coordinates, time stamps, speeds and course angles during vehicle accidents; Gsensor module for collecting triaxial acceleration data to evaluate collision direction and intensity; The OBU module is used for reading vehicle CAN bus data, including engine speed, brake state, accelerator opening, safety belt state, turn signal lamp state, ABS/EPS activation state and vehicle fault code; The video acquisition and encoding module is used for recording video data before and after an accident; the remote communication module is used for transmitting the acquired data to the cloud platform; and the local storage unit is used for caching the data and supplementing the data after the network is restored.
3. The vehicle insurance loss assessment system based on the vehicle-mounted electronic device and the AI glasses according to claim 2, wherein the cloud platform comprises: The data fusion engine is used for integrating high-precision positioning data, gsensor data, vehicle CAN bus data, vehicle-mounted equipment video data and AI glasses video data to generate an accident reconstruction model; the damage evaluation engine classifies and quantifies damage parts based on a deep learning algorithm, and comprises damage area, damage depth and part type identification; and the damage assessment decision engine is used for combining the damage assessment result, the accident dynamic data and the historical claim settlement database to output a maintenance scheme, cost estimation and damage assessment conclusion.
4. The vehicle insurance damage assessment system based on the vehicle-mounted electronic equipment and the AI glasses according to claim 3, wherein the AI glasses comprises: the high-definition camera and the depth sensor are used for capturing high-definition images of the damaged parts outside the vehicle and 3D point cloud data in real time; the AR display module is used for displaying damage evaluation results, maintenance suggestions and damage assessment codes through the augmented reality interactive interface; and the communication link is used for carrying out data exchange with the cloud platform.
5. The vehicle insurance loss system based on the vehicle-mounted electronic device and the AI glasses according to claim 4, wherein the data fusion engine comprises, in operation: S11, receiving high-precision positioning data, gsensor data, vehicle CAN bus data and vehicle-mounted equipment video data from vehicle-mounted electronic equipment, video data from AI glasses and 3D point cloud data, wherein the high-precision positioning data comprises accurate position coordinates, time stamps, speed and course angles of an accident moment, the Gsensor data comprises triaxial acceleration information to evaluate collision direction and energy, the CAN bus data comprises a braking state, accelerator opening and a vehicle fault code, and the video data covers whole process records before and after the accident; s12, carrying out time synchronization and alignment on the received data, ensuring that GNSS positioning data, gsensor acceleration data, CAN bus data and video streams are consistent on a time axis, and eliminating time sequence errors; s13, fusing GNSS positioning data and Gsensor acceleration data, and calculating kinetic energy at the moment of collision and a vehicle motion track, wherein the kinetic energy is estimated based on an acceleration peak value and mass, and the motion track is reconstructed through integrating the acceleration and position data; s14, analyzing the operation behaviors of a driver by combining CAN bus data, eliminating the interference of manual operation and assisting in restoring accident responsibility; S15, integrating video data of vehicle-mounted equipment and AI (advanced technology attachment) glasses, extracting vehicle attitude change and environmental characteristics through a computer vision algorithm, and performing cross verification with sensor data to generate accident process simulation animation, wherein the animation dynamically displays collision sequences, vehicle displacement and damage occurrence points; S16, outputting an accident reconstruction model, wherein the accident reconstruction model comprises a collision energy quantized value, a motion trail graph, a simulation animation and a data consistency report, and is used for subsequent damage assessment and damage assessment decision.
6. The vehicle insurance damage assessment system based on the vehicle-mounted electronic device and the AI glasses according to claim 5, wherein the damage assessment engine comprises, in operation: S21, multi-source data receiving and preprocessing, namely receiving high-definition image data and 3D point cloud data from AI glasses and video data from vehicle-mounted electronic equipment, denoising, illumination correction and scale normalization processing are carried out on the image data, registering and filtering are carried out on the 3D point cloud data, and acquisition environment interference is eliminated; s22, detecting and segmenting a damaged area, carrying out damaged area identification on the preprocessed image by using a deep learning model, positioning a damaged boundary, extracting a pixel-level mask of a damaged part, and simultaneously, carrying out spatial clustering on point cloud data by using a 3D point cloud segmentation network to identify solid geometric characteristics of the damaged area; s23, quantitatively calculating damage parameters, and calculating damage area, damage depth, damage length and damage morphological characteristics based on the segmentation result; S24, identifying and matching the type of the part, combining a vehicle model database, identifying the part to which the damage belongs through a convolutional neural network classifier, and associating the part number, the material property and the part importance level; s25, classifying the injury grades, and classifying the injury severity into three grades of slight, moderate and severe according to the quantitative parameters; S26, multi-mode data cross validation, namely logically comparing a visual evaluation result with vehicle-mounted sensor data to eliminate misidentification caused by illumination, shadow or angle; and S27, generating a structured damage report, and outputting a standardized report containing damage types, damage parameters, component information, damage levels and confidence degrees for the damage assessment decision engine to call.
7. The vehicle insurance damage assessment system based on the vehicle-mounted electronic device and the AI glasses according to claim 6, wherein the damage assessment decision engine comprises, in operation: S31, receiving a structured damage report output by a damage evaluation engine, accident dynamic data generated by a data fusion engine and maintenance cases and market price data in a historical claim settlement database; s32, matching a predefined maintenance rule base based on the part type and the damage level in the damage report to generate a maintenance scheme; S33, calculating material cost and labor hour cost by combining a maintenance scheme, a vehicle model database and a real-time market maintenance price database, and estimating cost; S34, if the accident dynamic data indicate that the collision energy exceeds a safety threshold value and the damage report identifies the damage of the key structural member, automatically triggering a high-risk vehicle mark, and generating a special detection flow; S35, generating a standardized damage assessment report, wherein the damage assessment report comprises a maintenance scheme list, a cost detail list, damage assessment codes and high risk prompts; And S36, displaying the damage assessment conclusion to a surveyor in real time through an AR interface of the AI glasses, supporting manual correction, and submitting the damage assessment conclusion to an insurance company claim settlement system.
8. The vehicle insurance damage assessment system based on the vehicle-mounted electronic device and the AI glasses according to claim 7, wherein when a predefined repair rule base is matched based on the type of the component and the damage level in the damage report, the predefined repair rule base includes the following rules when generating a repair scheme: A11, if the damage grade is slight and the part is an appearance part, generating a repairing scheme; A12, if the damage level is any one of the two conditions that the medium level and the component is a key structural member, generating a replacement scheme; A13, if the hidden damage is detected, adding a depth detection suggestion.
9. The vehicle insurance loss assessment system based on the vehicle-mounted electronic equipment and the AI glasses according to claim 8, wherein when the material cost and the labor cost are calculated by combining a maintenance scheme, a vehicle model database and a real-time market maintenance price database, and the cost is estimated, the specific rules are as follows: a21, dynamically matching the material fee based on the part number and the supplier quotation; a22, the labor hour cost is weighted and adjusted according to the maintenance complexity; a23, total cost is added to accident responsibility proportion.
10. The vehicle insurance damage assessment method based on the vehicle-mounted electronic equipment and the AI glasses is characterized by comprising the following steps of: s41, automatically recording dynamic data, including sensor data and video data, of the vehicle-mounted electronic equipment when an accident occurs, and transmitting the dynamic data to a cloud platform through a remote communication module; s42, a surveyor wears AI glasses to carry out panoramic scanning on the vehicle, and high-definition images and 3D point cloud data of the damaged part are collected; S43, analyzing vehicle-mounted equipment data by the cloud platform, calculating collision energy and a vehicle motion track, and generating an accident process simulation model; s44, carrying out damage assessment by the cloud platform based on data acquired by the AI glasses, and detecting damage types, damage areas and corresponding components; S45, the cloud platform cross-verifies the dynamic data of the vehicle accident process and the visual injury evaluation data to eliminate misjudgment; S46, generating an assessment code according to the damage severity and the importance of the parts; S47, combining the damage code, the market maintenance price database and the accident responsibility proportion, and outputting a maintenance expense list; If a potential safety hazard is detected, a high risk vehicle signature is triggered and depth detection is suggested S48.

Description

Vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses Technical Field The invention relates to the technical field of AI (advanced technology interface), in particular to a vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses. Background The vehicle insurance investigation damage assessment technology is a core link of the insurance industry for processing traffic accident claims, and relates to the evaluation of damage degree of accident vehicles so as to determine maintenance cost and insurance responsibility. Traditionally, this process relies on insurance companies dispatching professional surveyors to attend accident sites, manually inspecting the vehicle for damage, and making comprehensive decisions in conjunction with the site environment. The subject belongs to the technical field of vehicle-mounted electronic equipment and vehicle insurance investigation, and along with technological development, automatic elements are gradually introduced, but the whole is still mainly dominated by manpower, and the accuracy and efficiency of damage assessment are improved. The prior art mainly adopts a manual on-site investigation scheme, namely, an inspector inspects the damaged part by visual observation, takes a photo record and judges the damage degree based on personal experience. The method has the obvious defects of low efficiency, high subjectivity, inconsistent damage assessment results caused by the difference of judgment standards of different surveyors, easy disputes, incomplete data, difficulty in comprehensively restoring accident processes due to the lack of dynamic data such as key information of speed, acceleration, collision force and the like before and after the vehicle accident only by means of static photos and field observation, high cost, great manpower resources required for manual survey and difficulty in controlling operation cost in remote areas or high-frequency accident scenes because the manual investigation consumes a long time, especially in a multi-vehicle collision or complex accident. While enterprises have attempted to automate with image recognition or sensor data in recent years, these schemes rely on a single data source, such as an image or sensor, and fail to achieve fusion of multi-source data, resulting in limited accuracy of evaluation. In order to solve the problems, the invention provides a vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses. Disclosure of Invention In order to overcome the problems in the background art, the invention provides a vehicle insurance damage assessment system and method based on vehicle-mounted electronic equipment and AI glasses. The technical scheme of the invention is that the vehicle insurance damage assessment system based on the vehicle-mounted electronic equipment and the AI glasses comprises the vehicle-mounted electronic equipment, a cloud platform and the AI glasses, and is characterized in that: the vehicle-mounted electronic equipment is arranged in the vehicle and used for collecting dynamic data related to accidents; The cloud platform is in communication connection with the vehicle-mounted electronic equipment and the AI glasses and is used for data fusion, damage assessment and damage assessment decision; AI glasses are worn by surveyors for on-site vision acquisition and interactive display. Preferably, the in-vehicle electronic device includes: The high-precision GNSS positioning module is used for acquiring accurate position coordinates, time stamps, speeds and course angles during vehicle accidents; Gsensor module for collecting triaxial acceleration data to evaluate collision direction and intensity; The OBU module is used for reading vehicle CAN bus data, including engine speed, brake state, accelerator opening, safety belt state, turn signal lamp state, ABS/EPS activation state and vehicle fault code; The video acquisition and encoding module is used for recording video data before and after an accident; the remote communication module is used for transmitting the acquired data to the cloud platform; and the local storage unit is used for caching the data and supplementing the data after the network is restored. Preferably, the cloud platform includes: The data fusion engine is used for integrating high-precision positioning data, gsensor data, vehicle CAN bus data, vehicle-mounted equipment video data and AI glasses video data to generate an accident reconstruction model; the damage evaluation engine classifies and quantifies damage parts based on a deep learning algorithm, and comprises damage area, damage depth and part type identification; and the damage assessment decision engine is used for combining the damage assessment result, the accident dynamic data and the historical claim settlement database to output a maintenance sc