The detection and enforcement of over-speeding regulations remain a significant challenge due to limitations in existing vehicle speed estimation techniques. This study addresses this issue by developing a high-accuracy vehicle speed estimation system that incorporates a novel approach for license plate region extraction and over-speeding detection. The methodology employed involves the use of background subtraction methods to estimate vehicle speed, combined with a proprietary algorithm for vehicle plate region extraction. This information is then processed by a reporting system that identifies over-speeding vehicles. Results from extensive testing reveal a mean absolute error of 0.93 and a root mean square error of 1.40 in speed estimation, demonstrating high accuracy and precision of the developed system. Additionally, the mean absolute percentage error of 3.38% further substantiates the effectiveness of the system, leading to an overall accuracy of 96.62% in speed estimation. This advancement in traffic management technology has the potential to improve road safety, reduce traffic violations, and contribute to more efficient and streamlined urban planning.
Danladi et al. (2025) studied this question.