基于HSV颜色空间检测与Gabor筛选器的阴影检测

Shadow detection based on HSV color space characteristic and Gabor screener

针对视频分析中难以完全将前景(FG)和运动阴影正确分离,提出一种基于阴影HSV颜色空间特性与Gabor筛选器的阴影分割方法。首先,采用一种基于复杂背景(BG)的运动目标检测方法提取出运动目标;其次,采用基于HSV颜色空间阴影特性初步判定阴影区域;最后,设计基于感兴趣区域(ROI,region of interest)的Gabor筛选器对初步判定后的阴影区域进行筛选,从而检测出阴影。对不同光照和环境条件下的视频序列进行测试结果表明,方法效果好,阴影检测率高,可应用于智能视频监控的目标检测。

It is difficult to segment moving shadow from foreground completely in the video analysis.A method of shadow segmentation based on shadow HSV color space feature and Gabor screener is proposed.Firstly,a detection method of moving target based on complex background is used to extract the moving target from the video sequence.Then the shadow HSV color space feature is considered initially to judge the shadow region.Further the shadow region is judged by adopting the Gabor screener on the ROI（region of interest）.In order to evaluate the performance of the proposed way,the video sequence from different light and environmental conditions is detected.The result shows that this method is effective,and has higher shadow detection rate,it can be applied into the target detection of the intelligent video surveillance.