Vehicle Damage Assessment System
via 3D Model Retrieval Techniques
Abstract
Nowadays, motor insurance is an insurance that all road users must have as a guarantee
to protect life and property in the event of an accident from using a road vehicle. However, motor
insurance often encounters problems in the process of vehicle damage evaluation to settle
insurance compensation payments for the insured.
The developer therefore proposed a system for assessing car damage by using a 3D model
retrieval method by converting the pre-insurance 3D model and the accident 3D model into a 2D
multi-view image. The two models were compared to determine the damage point of the car
and then classified it into the severity of the damage. The damage assessment system has an
accuracy of 91.25% based on training with 8 3D models of the car model. and tested with 5 3D
car models
The benefits of developing this system are This system will help in the initial damage
assessment. Helps reduce costs incurred from damage assessment It can also be further
developed for use in bone comparison to create artificial bone in medicine. or to develop to
compare the model of the car model Packaging and other objects such as phones, shoes, laptops,
or objects that can be 3D modeled
The website template was borrowed from Michaël Gharbi.