Optimized Car Damaged Detection using CNN and Object Detection Model

Car damage detection systems have become very significant in various daily activities, like the insurance claims process, accident reporting systems, or a documented write up of the various damages a vehicle in a car garage has when it enters and exits the garage. In this paper the authors present a machine learning algorithm based on a two different model approach, a convolutional neural network (CNN) used to detect various damages in a car and classify them into Scratches, Dents, Glass Shattered and Smash using an object detection model. The CNN model filters out the undamaged images, passing only the image with damages to the object detection model, to augment the model's performance. The purpose of this research project is to bring maximum accuracy with the use of CNN model. The research project was conducted using TensorFlow module for pretrained framework atop to build our object detection model.