Vehicle Damage Assessment Image Recognition Using Part List Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current vehicle damage assessment image recognition technologies face challenges in accurately identifying damaged automotive parts due to variations in vehicle configurations, leading to high implementation costs and long training periods, with limited accuracy resulting from reliance on image recognition algorithms alone.

Innovation Solution

The method involves obtaining a vehicle identification code to access a comprehensive automotive part list, which is then matched with the identified damaged part to determine the correct automotive part and its serial number, improving recognition accuracy and reducing costs by leveraging existing configuration information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If purely relying on model algorithm for image recognition, then implementation cost is reduced, but recognition accuracy is limited

Engineering Contradiction:
Improveimplementation costVSAvoidrecognition accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent introduces vehicle configuration information as an intermediary element that mediates between the image recognition algorithm and the final identification result. The configuration information acts as a reference database that guides and corrects the algorithm's output, enabling accurate part identification without requiring expensive and time-consuming retraining of the image recognition model itself.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If collecting as many appearance image data as possible for training, then recognition precision is ensured, but training period and implementation costs increase

Engineering Contradiction:
Improverecognition precisionVSAvoidtraining period
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-collecting and organizing vehicle configuration information (part names, serial numbers, specifications) into a structured database before the actual damage assessment process. This preliminary preparation of reference data eliminates the need for extensive image training, as the system can directly match recognized parts against the pre-prepared configuration information.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The vehicle configuration information serves as an intermediary that bridges the gap between limited training data and high recognition precision. Instead of relying solely on大量 training images, the system uses the configuration database as a reference to verify and refine algorithm results, achieving high precision without extensive training.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Adaptability or versatility

If using more comprehensive training data covering various vehicle configurations, then adaptability improves, but data collection and processing complexity increases

Engineering Contradiction:
Improveadaptability to vehicle configurationsVSAvoiddata collection and processing complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts the essential configuration information (part names, serial numbers, specifications) from the complex vehicle data and separates it into a dedicated reference database. This extraction approach allows the system to handle various vehicle configurations without increasing the complexity of the image recognition algorithm itself, as the configuration variations are managed in the separate reference database rather than requiring complex algorithmic processing.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3617949B1Method and apparatus for improving vehicle loss assessment image identification result, and server
Publication Date: 2024.07.24 ADVANCED NEW TECHNOLOGIES CO LTD
  • EP3617949B1 patent drawingFigure 1
  • EP3617949B1 patent drawingFigure 2~3
  • EP3617949B1 patent drawingFigure 4~5

AI summary

Embodiments of this specification disclose a method and an apparatus for improving a recognition result based on a vehicle damage assessment image, and a server. The method includes: obtaining a preliminary damaged automotive part of a vehicle, where the preliminary damaged automotive part includes a damaged automotive part of the vehicle that is obtained by recognizing a damage assessment image by using a preset image recognition algorithm; obtaining an automotive part list of the vehicle, where the automotive part list includes a plurality of automotive part identification serial numbers corresponding to automotive part data; matching the preliminary damaged automotive part with the automotive part list, to determine, in the automotive part list, an automotive part corresponding to the preliminary damaged automotive part; and outputting an automotive part identification serial number of the matched automotive part. Combining automotive part list information and recognition of a vehicle damage assessment image can improve the accuracy of damaged automotive part recognition result based on the damage assessment image, and significantly reduce additional learning costs and the period of an image recognition algorithm/module.