Chapter 13 - Smart automated system for classification of emergency heavy vehicles and traffic light controlling

Abstract

In this work, enhanced vehicle classification and traffic controlling systems are proposed to replace the traditional methods of road traffic control systems. The proposed system comprises an intelligent model in which traffic lights are connected to the remote server through the Internet. The primary objective of the proposed system is to estimate the real-time traffic congestion on the roads based on enhanced image-processing techniques such as deep neural network models and background subtraction methods. The next objective is to prototype a model for controlling traffic light signals. For this purpose, the real-time traffic information is gathered, processed, and analyzed based on video images captured by smart video cameras on the traffic light poles. The traffic light density is estimated and used to decide which side of the road to clear first to minimize standing time, air pollution, noise pollution, and oil consumption. This approach to a traffic light system includes a system that can classify between emergency and nonemergency vehicles and clear the roads that have emergency vehicles first. A smart traffic light can be used for many things other than those mentioned here; it can be used to read number plates and check with police databases to continuously scan for stolen vehicles, to detect the speed of vehicles, and even to issue notifications to owners of vehicles using number plate recognition.