面向室外视频监控的背景重构算法  

A Background Reconstruction Algorithm for Outdoor Video Surveillance

作　　者：郑锦[1] 李波[1] 

机构地区：[1]北京航空航天大学计算机学院数字媒体室,北京100191

出　　处：《电子学报》2009年第8期1854-1859,共6页Acta Electronica Sinica

基　　金：国家863高技术研究发展计划(No.2009AA01Z316);国家自然科学基金(No.60505007)

摘　　要：针对室外视频监控环境复杂,运动对象检测准确性较低,提出一种用于背景减运动对象检测的背景重构算法.综合考虑像素点亮度的稳定状态持续时间和出现频率,定义加权亮度直方图确定量化区间的背景概率,根据背景概率大小和分布,结合基于亮度空间相关性的修正实现背景重构.实验结果表明,重构的背景有效地克服了训练阶段运动对象的干扰,用于运动对象检测能够适应背景扰动、摄像机轻微晃动等复杂情况,提高检测的准确性.The accuracy of motion objects detection is low for outdoor complicated visual surveillance,so a background re- construction algorithm used in background subtraction motion objects detection is proposed. The algorithm integrates the stab＇tlization duration with appearance frequency, and uses weighted intensity histogram to confirm the background probability of quantity interval. Based on the probability values and distributing, it uses the intensity space correlation to amend the background, and realizes the background reconstruction. The experimental results show that the reconstructed background conquers the influence of motion objects in waining phase, and in motion objects detection application, it is applicable in the situation existing background fluctuation and small camera displacements etc, and it improves the accuracy of detection.

关 键 词：运动对象检测 背景重构 加权亮度直方图 背景概率 空间相关性 

分 类 号：TN919.8[电子电信—通信与信息系统]