Automated Traffic Law Enforcement System

Road accidents and traffic offenses hinder growth and cost Sri Lanka money. From 2016 to June 2023, 223,000 accidents killed over 20,000 people, 8 every day. Road fatalities are now over 120 per million, greater than the US and Japan. Rapid motorization without infrastructure growth, lax enforcement, absence of speeding and drunk driving punishments, and defective violation reporting systems are important issues. An Automated red Violation Detection, Reporting, and Fine System uses computer vision, deep learning, and IoT to identify speeding, lane breaches, and red-light disobedience via video cameras. It would immediately report offenses and publicly fine without traffic police. Vehicle detection, speed identification, lane infraction recognition, and traffic light classification are accurate under real-world scenarios thanks to Deep Learning models. Over 90% speed violation detection accuracy in diverse weather is achieved with extensive data augmentation. Through 24/7 monitoring and transparency, automated systems deter noncompliance, reduce accidents, and improve road safety. Results show that traffic video insights can be used to construct intelligent law enforcement systems.