Date of Award

5-2025

Document Type

Thesis

Degree Name

Master of Science in Information Systems and Technology

Department

Information and Decision Sciences

First Reader/Committee Chair

Benjamin Becerra

Abstract

Growing vehicle usage has resulted in a notable increase in road accidents, so it is imperative to have effective systems for identifying and evaluating vehicle damage. This work aims to create a computer vision and deep learning-based automated car damage detection system. This project's main goal is to develop a model that, using visual cues, can categorize car photos as either damaged or undamaged.

The algorithm operates in two steps: first, determining whether the picture features an automobile; then, it classifies the state of the car—damaged or undamaged. We thus employ the InceptionV3 model for damage classification and the MobileNet SSD (Single Shot Multibox Detector) for vehicle detection to do this.

The image contains an automobile detected using a pre-trained object detection model called MobileNet SSD. Downloaded automatically during the procedure, the pre-trained set of weights injected into the model is once a car is spotted. The image is then fed to the InceptionV3 model, which has been optimized specifically for damage detection. Using a dataset of automotive photos, where every image is tagged as either damaged or undamaged, the InceptionV3 model was trained. With a 94% accuracy, the algorithm forecasts the car's damage status, proving great real-time predictive efficiency.

Training uses a dataset of manually labeled car photos for damage and undamaged categories. The model detects vehicles using the COCO dataset labels; the car is classified as class 7 there. The algorithm responds with an "invalid image" when no car is detected in the image, handling random images with unknown backgrounds—such as objects or landscapes.

The project also combines other methods, including data augmentation to lower overfitting and enhance the generalization of the model. A major benefit of the model is its capacity to manage changes in illumination, angles, and image quality, therefore qualifying it for use in practical environments.

This method may be expanded to assist many sectors, including automotive insurance, where it can automate the damage assessment process, saving the time and effort needed by human assessors. To speed damage assessment, it might also be included in mobile apps or used in systems handling insurance claims.

The effective execution of this research shows the possibilities of deep learning for automating automotive damage detection. Expanding the dataset, increasing model accuracy, and investigating real-time implementation to handle video streams or live photos for instantaneous predictions will be the main priorities of the following work.

Download

  Contact Author

Share

COinS