Tint Detection Using Image Analysis
Year: 2025
Type: article
Abstract: This paper introduces an automated Tint Detection System designed to meet regulatory requirements for vehicle window tinting through advanced image analysis techniques. Excessive tinting on vehicle windows can impair visibility, affecting road safety and law enforcement’s ability to monitor vehicles. Traditional manual inspections are time-intensive, costly, and prone to error. This project proposes a system that automates tint detection in real-time using a modular pipeline incorporating YOLOv5 for vehicle detection, U-Net for window segmentation, and a Convolutional Neural Network (CNN) for Visual Light Transmission (VLT) analysis. Our system facilitates efficient, automated detection of tint levels, supporting law enforcement in ensuring compliance with minimal manual intervention.
Authors: Yokesh Babu S, Vansh Harkut, Himanshu Sharma, Vishal
Institution: Vellore Institute of Technology University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4408623869