Vehicle Flow Detection Based on Machine Vision

摘要

Vehicle flow detection is the key technology of intelligence traffic system. In order to overcome the shortcomings of traditional detection method which with low detection precision and is difficult to install and maintenance, a method, based on machine vision, which combines adjacent frame difference and background frame difference is proposed, besides, another method of vehicles count by vertical projection is also proposed. The simulation results show that the proposed vehicle detection algorithm not only can achieve high detection accuracy, but also is very simple and fast.