Title – Intelligent Traffic Signal System using Classification of Vehicular Traffic Using Lazy and Function Family Data Mining Classifiers

Author(s) – Nikita Bhavsar, Aditya Bhatkhande, Ruchira More, Shubham Vasaikar, Sudeep D. Thepade

Country – India

Abstract – This paper witnesses a novel Vehicle Traffic signal switching methodology based on the estimated density of the traffic over a road. Depending on the density of the traffic the signal stretch would be altered dynamically. This paper takes yet one more progressive step in the construction of smart city. The method used here is a simple logic that extracts feature vectors which have both reduced time complexity and computational complexity. Feature vectors such as row mean, column mean diagonal mean and their permutations are calculated on the subtracted image and passed to the system for deciding the stretch of the signal. The key to the system is the dataset passed in the form of frames by a CCTV camera mounted on the signal. The system outputs the classified frames as low, moderate or high. The RBF Classifier has given the best accuracy along with the Column Mean feature vector.

Keywords – Traffic density, row mean, column mean, diagonal mean, classification, vehicle detection, data mining, Lazy, Function.

Full Text – Download PDF RJ020201