摘要
从煤矿工业电视监控系统的现状出发,提出了一种煤矿安全异常检测算法,采用有条件背景替换的思路进行检测,能够避免光照、噪声等因素的影响,满足煤矿实时监控的需要。同时,提出了一种评价视频异常检测算法效率的方法,并构造了检测标准视频序列。
An algorithm of colliery security anomaly-detection is presented from the present condition in coal industry video monitoring system. The thinking of substituent with conditional background is employed to test. It can avoid the influences of illumunation and noise etc, and meet the real-time monitoring of mine. And an evaluation methodology evaluating anomaly-detection efficiency is given. Detection result shows this algorithm is suitable for real system.
出处
《传感器与微系统》
CSCD
北大核心
2008年第4期66-67,73,共3页
Transducer and Microsystem Technologies
基金
辽宁省科技攻关计划资助项目(2006220019200422900)
关键词
煤矿
异常检测
评价方法
视觉监视
mine
anomalies detection
evaluation methodology
visual monitor
