基于SOPC实时运动目标检测与跟踪系统.pdfVIP

第 25卷 第 4期 重 庆 理 工大学 学 报(自然科学) 20 年 4月 Vol. 25 No. 4 Journal of Chongqing University of Technology( Natural Science) Apr. 20 SOPC 2 李月静 , 谢维成 , 石 一兴 , 陈昌敏 ( . , 6 0039; 2. , 6 0064) : 设计了一种以 FPGA 核心处理器件的实时运动目标检测与跟踪系统, 采用基于 累积差分更新的背景减除法检测运动目标, 卡尔曼滤波和直方图匹配算法对目标进行跟踪实 验结果表明, 该系统能对静态背景视频序列中的运动目标实时有效地进行检测和跟踪 : 视频序列; 检测与跟踪; SOPC; 背景减除; 卡尔曼滤波 : TP39 : A : 674- 8425( 20 ) 04- 009 - 06 R ea-l tim e M oving Object Detection and T rack ing System based on SOPC 2 LIYue-jing , X I EW e-i cheng , SH IY -i xing, CHEN Chang-m in ( . School of E lectrical Engineering and Inform ation, X ihua University, Chengdu 6 0039, China; 2. School of E lectron ics and Information, Sichuan Un iversity, Chengdu 6 0064, Ch ina) Abs tract: The rea-l tmi e perform ance has been increasingly requ ired w ith the progress of intelligent video surveillance system, but relevant algorithm is increasing comp lex. To realize the rea-l tmi emov- ing object detection and tracking, it awl ays utilizes the system on a programm able ch ip ( SOPC) w ith h igh speed and concurrent flow processing ab ility, wh ich has aw ide app lication prospect. Th is article designed a rea-l tmi emoving object detection and track ing system adopted by FPGA as core processor. The system used background subtraction method based on accumu lative d ifference updating for detec- ting moving object. The Kalman filter and h istogram matching algorithm have been used for object tracking. The expermi ental resu lts show that the system can