Smart Moving Vehicle Counting

Abstract

The moving vehicle detection and counting in video is a substantial field of computer vision. In traffic observation, the number of vehicles detected might be used to survey or control the traffic signal, therefore several counting methods have been suggested. The proposed system used static camera to capture AVI video format, several main steps that applied to detecting and counting the moving vehicle in the input video. This steps are: frames extraction, frames preprocessing, features extraction, vehicle detection and vehicle counting in one side road and in two side road. The experiments showed reliable and efficient results of the moving vehicles in any traffic sense within 99.25% accuracy of one side road counting processes and within 95.23% accuracy of two side road processes.