一种基于彩色特征的交通流量统计方法  

A Vehicle Detection Method Based on Color Information

作　　者：刘震 李宇[2] 刘彦飞[3] 

机构地区：[1]重庆药品交易所,重庆400010 [2]重庆市公安局科技信息化处,重庆401147 [3]重庆市公安局大渡口分局,重庆400084

出　　处：《电子质量》2013年第3期73-76,共4页Electronics Quality

基　　金：重庆市科学技术委员会科技创新服务类项目资助(编号:cstc2012cx-rkxA0246)

摘　　要：该文基于对传统帧差法的改进,运用二次帧差法来进行车辆特征信息的提取,并运用彩色信息和车辆偏移量相结合的方法进行计数,提出一种环境自适应能力强、适合于运动车辆检测的新方法。大量实验证明,该方法可正确检测车辆,实现车流量统计。该检测算法采用颜色特征和车辆偏移量进行判断,提高了检测精度,大大改善了车辆检测效果,为实现交通管理的自动化奠定基础。Based on the improvement of the traditional frame-difference method,by the use of secondary frame difference method to the extraction of the vehicle characteristics and the method of combining color information and vehicle offset,we propose an adaptive and suitable moving vehicle detection method.A large number of experiments show that the method can correctly detect vehicles,achieve traffic statistics. The detection algorithm uses color features and vehicle offset to detect.It can improve the detection ac-curacy,and greatly improve the effect of vehicle detection,lay the foundation for the realization of the au-tomation of the traffic management.

关 键 词：车辆特征信息 二次帧差法 彩色特征 车辆检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]