摘要
本文提出一种应用于智能交通监控系统的车辆和行人的检测和识别方法。首先结合帧间差分,对交通监控场景的视频图像序列建立动态背景模型,然后用背景消除法对监控视频中的运动目标进行有效检测,提取出运动目标的轮廓。最后采用支持向量机,对检测出运动目标进行快速识别。实验结果表明,该方法能够快速准确地在监控视频中对运动的车辆和行人进行检测和分类识别,对快速交通通道中非法行人入侵进行自动报警。
In this paper, an approach to detect and recognize the moving vehicles and pedestrian that applied in intelligent traffic surveillance is proposed. Firstly, a dynamic background model is created with frame differencing for video sequence images of traffic surveillance scene. Then the moving object in the video can be detected effectively with background subtraction. Finally, SVM is used to recognize the moving objects. The experiment results show this technique can effectively detect and recognize moving vehicles and pedestrian in the surveillance video in time.
出处
《电子测量技术》
2007年第1期16-17,71,共3页
Electronic Measurement Technology
关键词
智能交通监控
目标检测
目标识别
支持向量机
intelligent traffic surveillance
object detection
object recognition
SVM
