Abstract
This research introduces a machine learning approach to detect the speed of vehicles. Our proposed system utilizes computer vision algorithms to track and identify moving vehicles in time. It then employs a trained machine learning model to estimate their speed based on the collected data. The methodology relies on a network (CNN) architecture, which is trained using a substantial dataset of vehicle images and corresponding speed measurements. Our system exhibits accuracy and reliability in estimating speeds across test scenarios encompassing different types of vehicles and lighting conditions. An optimum vehicle count is recorded with heavy vehicles in place as compared to other vehicle types. A mean response delay of 1.25 seconds and a RMSE value of 0.05 is observed with less road traffic in place. The suggested technology holds applications, in transportation systems, traffic monitoring and enhancing road safety.