摘要
为检测视频监控系统中运动目标的阴影,提出了一种基于光照不变特征c1c2c3的检测算法.该算法首先根据前景像素和处于相同位置的背景像素的c1c2c3之比定义坐标系,并将阴影检测视为该坐标系内阴影和运动物体的分类问题;然后,采用阴影像素在该坐标系中的分布特性构建1个椭球,并将处在椭球内部的像素判别为阴影像素;最后,根据阴影区域的几何信息进行后处理.实验表明,该算法能有效地检测不同场景、不同运动物体的阴影,并能适应光照条件的变化.
To detect moving shadow in intelligent video surveillance system,a method using illumination invariants c1c2c3 is presented.Firstly,the coordinate systems are defined on the basis of c1c2c3 ratios of foreground and background pixels at the same position.Secondly,the moving shadow detection can be treated as a classification problem of shadows and moving objects in abovementioned coordinate systems.An ellipsoid is further constructed according to the scattering feature of shadow pixels.All the pixels inside the ellipsoid are extracted as shadows. Finally, the postprocessing is performed by exploiting the geometrical property of shadows. Experiments show that the proposed algorithm is robust and effective in detecting shadows for a variety of scenes and moving objects; moreover, it is suited for varying illumination conditions.
出处
《北京邮电大学学报》
EI
CAS
CSCD
北大核心
2009年第5期109-113,共5页
Journal of Beijing University of Posts and Telecommunications
基金
国家自然科学基金项目(60743007
60872050)
关键词
模式识别
运动阴影检测
智能视频监控
pattern recognition
moving shadow detection
intelligent video surveillance