基于DSP的视频车辆实时检测系统的实现

被引:4
作者
鲁一杰
赵攀
彭强
江永全
机构
[1] 西南交通大学信息科学与技术学院
关键词
数字信号处理器; 大规模集成参考框架; 车辆检测; 背景提取; 阴影检测; 运动检测;
D O I
暂无
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
描述了基于TI公司TMS320DM642型DSP平台的视频车辆实时检测系统的实现,针对DSP系统在运算和内存方面的局限性,提出了一种线性时间复杂度下的矩形区域多目标搜索算法,解决了DSP栈深度的困难。应用高性能RF5参考框架进行算法构建,通过改进的阴影检测和运动目标检测算法,从而实现视频车辆的精确实时跟踪。实验证实了算法的有效性和处理的实时性,实现了车辆实时检测的微型化。
引用
收藏
页码:757 / 759+765 +765
页数:4
相关论文
共 10 条
[1]  
Vision-basedvehicle classification. GUPTE S,MASOUD O,PAPANIKOLOPOULOS NP. Proc of IEEE Intelligent TransportationSystems Conference . 2000
[2]  
Computer vision algorithms for intersection monitoring. Veeraraghavan,H,Masoud,O,Papanikolopoulos,N. IEEE Trans Intell Transport Syst . 2003
[3]  
Detection and classification of vehicles. GUPTE S,MASOUD O,MARTIN R,et al. IEEE Transaction on Intelligent Transportation Systems . 2002
[4]  
Introduction to al-gorithms. CORMEN TH,LEISERSON C E,RIVESTR L. . 2001
[5]  
The sakbot system for moving object detection and tracking. Cucchiara R,,Grana C. Video-Based Surveillance Systems-Computer Vision and Distributed Processing . 2001
[6]  
Improving shadowsuppression in moving object detection with HSV color information. CUCCHIARA R,GRANA C,PICCARDI M,et al. Proc of IEEE Int’l Conference on Intelligent TransportationSystems . 2001
[7]  
Difference and accumulative difference pictures in dynamic scene analysis. Jain R. Image and Vision Computing . 1984
[8]  
Reference Frameworks for eXpress-DSP Software:RF5,An Extensive,High-Density System[OL]. Texas Instruments Inc. http://www.ti.com . 2003
[9]  
The Use of Computer Vision in Monitoring Weaving Sections. Osama Masoud,Nikolaos P.Papanikolopoulos,and Eil Kwon. IEEE Transactions on Intelligent Transportation Systems . 2001
[10]  
TMS320DM642 Technical Overview. Texas Instruments. http://focus.ti.com/docs/apps/catalog/resources/appnoteabstract.jhtml?abstractName=spru615 . 2002