Traffic Prediction for Intelligent Transportation Systems using Machine Learning

Over the past few decades, ITS have spiked an increasing research interest as a promising discipline for revolutionizing the transportation sector and solving common traffic and vehicle-related problems. ITS comprise a multitude of interconnected engineering feats that function as an entity for optimizing network-scale travel experiences from a technical, social, economic, and environmental aspect. Such optimizations necessitate the advancement of information and communication technologies, electronic sensors, control systems, and computers, which high-lights the data-driven nature of modern ITS. In this paper we design a system which uses machine learning algorithm using SVM, KNN and CNN algorithm which is a novel system which will provide intelligence to the current traffic control system present at a four-way junction. This ML technique is mainly aimed to replace the existing traffic light control system with artificial intelligence system. Nowadays most cities are equipped with CCTV cameras on the roads and the junctions, the basic idea is to collect the live video from the CCTV cameras and detect the number of vehicles on each lane and feed the data into another machine learning algorithm. according to the data of each lane changes into the light phase of the green signal. This system mainly aims to increase the traffic efficiency by increasing vehicle flow which will reduce waiting time for the vehicles. We are using HOG algorithm for feature extraction. In the implementation of the proposed architecture, we have achieved an accuracy of 86.34% for binary classification and 90.23% for multi-class classification