Machine learning-based smart speed gun for automatic ticket generation and identification of vehicles with pending tickets

Abstract

In Uganda, speed guns lack automatic ticket generation capability, forcing traffic officers to manually enter vehicle details, a process prone to bribery and ineffective enforcement. This contributes significantly to the high rate of accidents caused by over speeding. Meanwhile, existing speed cameras are limited in number, stationary, and easily evaded. Additionally, officers lack tools to identify vehicles with outstanding fines, relying on tiresome and inefficient manual checks. This project addresses these issues by developing a smart speed gun prototype integrating Machine Learning. The system utilizes a Raspberry Pi camera for real-time vehicle footage and employs a Machine Learning model for automatic license plate recognition. Speed calculations are performed using infrared sensors placed along vehicle paths. Upon detection of over speeding, the system issues a speeding ticket through SMS. Vehicles with existing fines trigger an alarm, sounded by a buzzer, to alert authorities. The prototype has undergone controlled testing, demonstrating capabilities in automatic ticket generation and identification of vehicles with pending road fines, potentially enhancing enforcement efficiency and reducing over speeding fatalities in Uganda.