Related Experiment Video
Updated: Aug 6, 2026
08:27
Image Recognition and Parameter Analysis of Concrete Vibration State Based on Support Vector Machine
Published on: January 5, 2024
An intelligent approach for automated vehicle damage classification
Shimaa Ouf1, Bassant Lotfy2, Soha Ahmed2
1Information Systems Department, Faculty of Commerce and Business Administration, Capital University (Formerly Helwan University), Helwan, Cairo, Egypt. Shimaaouf@commerce.helwan.edu.eg.
Scientific Reports
|July 20, 2026
Summary
This study introduces an AI framework for vehicle damage detection and repair cost estimation. Utilizing deep learning models like YOLOv8, it accurately identifies and classifies damage, improving efficiency and precision in assessments.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Automotive Engineering
Background:
- Manual vehicle damage assessment is time-consuming, resource-intensive, and prone to inconsistencies.
- Accurate damage classification and repair cost estimation are crucial for the automotive industry.
Purpose of the Study:
- To develop and evaluate a framework for automated vehicle damage detection, classification, and repair cost estimation.
- To compare the performance of YOLOv5 and YOLOv8 models for vehicle damage identification.
- To assess the efficacy of the XGBoost model in predicting repair costs.
Main Methods:
- Vehicle damage was detected and classified using YOLOv5 and YOLOv8 deep learning models.
- Repair cost estimation was performed using the XGBoost machine learning model.
- A custom dataset was created, featuring damage types, bounding box coordinates, and incorporated features.
Main Results:
- YOLOv8 demonstrated superior performance over YOLOv5 in damage detection and classification, achieving 87% precision (validation) and 90.7% mean absolute precision (test).
- The XGBoost model achieved a high R² score of 97.28% and a Mean Absolute Error (MAE) of 128.50 for repair cost prediction.
- The proposed framework significantly enhances the precision and efficiency of vehicle damage assessment.
Conclusions:
- The developed AI framework offers a precise and efficient solution for vehicle damage detection, classification, and cost estimation.
- Deep learning and machine learning models provide a viable alternative to traditional manual assessment methods.
- The framework shows strong potential for real-world application in the automotive sector.