摘要
本文阐述了一个基于ARM9的嵌入式智能视频入侵物识别系统。该系统利用入侵物捕获、跟踪算法对从摄像头采集的图形数据进行实时处理,从而对入侵物进行实时连续监控,识别其行为并据此判断其是否可疑,然后预测一个趋势并及时告警。该系统在实际场景中做了验证测试,证明其具有事前预防的作用,优于传统视频监控系统的事后调查取证。
This paper describes an embedded intelligent video system,which can identify invasion and based on ARM9.This system can real-time process graphics data（which is acquired by camera） with algorithms of invasion capturing and invasion tracking which is designed by the writer,in order to real-time and continuous monitor invasion,to identify its behavior and to judge with the identification whether the invasion is suspicious,and then,to forecast a trend and to warn in time.This system has the function of prevention,which is superior to the investigating and obtaining evidence after the event of traditional video surveillance system.Based on the test in actual scene,the writer proved that the system is feasible.
出处
《微计算机信息》
2011年第3期103-104,118,共3页
Control & Automation
关键词
智能视频
入侵物捕获
入侵物跟踪
Intelligent video
Invasion capturing
Invasion tracking
