Computer Vision Based Traffic Monitoring and Analyzing From On-Road Videos
Traffic monitoring and traffic analysis is much needed to ensure a modern and convenient traffic system. However, it is a very challenging task as the traffic condition is dynamic which makes it quite impossible to maintain the traffic through traditional way. Designing a smart traffic system is also inevitable for the big and busy cities. In this paper, we propose a vision based traffic monitoring system that will help to maintain the traffic system smartly. We also generate an analysis of the traffic for a certain period, which will be helpful to design a smart and feasible traffic system for a busy city. In the proposed method, we use Haar feature based Adaboost classifier to detect vehicles from a video. We also count the number of vehicles appeared in the video utilizing two virtual detection lines (VDL). Detecting and counting vehicles by proposed method will provide an easy and cost effective solution for fruitful and operative traffic monitoring system along with information to design an efficient traffic model.
