A Video-Based Method for Vehicle Counting Based on Motion and Color Features

Abstract

The vehicle detection and counting plays a crucial role in traffic control. With the development of intelligent transportation system (ITS), vehicle counting based on traffic video shows great importance and huge advantages compared with traditional technologies. However, the complexity and accuracy of many algorithms are often difficult to get a good balance. At the same time, the motion and color features of the vehicle are sometimes not fully utilized. Based on these, a video-based method for vehicle counting based on motion and color features was proposed. The vehicle will be detected based on motion and color features both to improve the accuracy of detection, and the counting information will be converted from detection information based on a horizontal line. The results show that the proposed method is effective on different videos.