Ife Journal of Information and Communication Technology

Abstract

Recognizing vehicle density is important in effectively controlling road traffic. Vehicle lanes with high density are given higher priority and more road access than vehicle lanes with lower densities to aid effective traffic control. The manual road traffic control method, which relies on human intuition for managing road traffic, is quite effective but prone to problems associated with manual systems, such as fatigue and unavailability. Other systems, such as traffic lights, sensors, RFID, etc., were introduced to solve the problems of the manual method. However, none of these methods is as effective in recognizing high-density vehicle lanes as the manual method. This research developed an intelligent high-density vehicle lane recognition system for traffic control. Digital cameras were used to obtain data on vehicular traffic at a road junction. After that, a convolutional neural network was then designed to recognize high-density vehicle lanes. This was then used to develop the intelligent traffic control system. The developed system obtained an accuracy of 92% in recognition of high-density vehicle lanes in a 4-way road junction. In conclusion, the intelligent system could mimic human intuition of recognizing high-density vehicle lanes for effective road traffic control.