摘要
目标识别是智能安防监控视频处理系统中重要内容,为了满足安防系统小型化实时性等应用需求,设计了一种基于FPGA平台的运动目标识别系统;该系统采用CMOS摄像头作为视频采集器,SDRAM作为视频流缓存及存储介质,利用FPGA可并行处理特点,采用流水线技术进行分模块化设计,对视频流进行灰度化和帧差法算法处理,并将结果传输至显示器终端,从而实现图像采集和目标识别实时跟踪和显示;测试结果表明,该系统在一定测距范围内可有效稳定地跟踪运动目标物体,可实时显示、小型化且功耗低,可进一步应用于安防领域中。
Target recognition and tracking is an important part of intelligent security monitoring video processing system.In order to satisfy requirements for real time miniaturization,a target recognition and tracking system based on FPGA platform is designed.The CMOS digital camera is used to collect original video data in the system,and the off-chip SDRAM is taken as the video stream buffer and storage medium.The image processing part is completed by FPGA,including gray processing,inter-frame difference method,and the results are transmitted to the display terminal to realize the video acquisition,target recognition,tracking and display.The test results show that the whole system can track the target object effectively and stably.It can be further applied to security the monitoring system.
作者
刘汝卿
李锋
蒋衍
朱精果
LIU Ruqing;LI Feng;JIANG Yan;ZHU Jingguo(Institute of Microelectronics of the Chinese Academy of Sciences,Beijing 100094,China)
出处
《计算机测量与控制》
2022年第4期56-59,共4页
Computer Measurement &Control
关键词
FPGA
实时目标检测
帧差法
FPGA
real-time target recognition and tracking
inter-frame difference method
