Video-Based Detection, Counting and Classification of Vehicles Using OpenCV

Abstract

The detection and counting of vehicles have significant role in an intelligent transportation system, especially for traffic management. Traffic issues has become a biggest problem where city planners facing for years. The detection of moving vehicles more accurately, several computer vision techniques, vehicle counting is done by using a virtual detection zone. Traffic analysis will calculate for the number of vehicles in an area per some arbitrary period of time and classify the vehicles. But the moving vehicles and their detection, tracking, and counting are very critical for flow of traffic in monitoring, planning, and controlling. By analyzing the recorded video of traffic flow sequence from a camera, a video- based solution technology is applied in combination with virtual detector and blob tracking technologies and YOLO are necessary. With this technology we applied Open CV for real time video applications. Such methods help us to detect, track, count and classify the moving vehicles.