一种实时检测遮挡阴影的算法

监控系统中,观测视场被遮挡,会导致显示的图像中存在遮挡阴影,影响对图像的进一步分析和处理.针对该问题,提出了一种实时检测遮挡阴影的算法.该算法基于图像序列的帧间运动特性对图像进行分区处理;利用图像子区域的灰度均值进行帧间预测,用来判断图像序列中被遮挡图像帧和可能的遮挡阴影区域;利用空域滤波器剔除误检区域,确定阴影区域.该算法实时性好,易于工程实现.

Shadow detection algorithm based movement feature

Shadowed area will appear in the displayed image when the field of view of a surveillance system is obscured, and it influences the analysis and further processing of the image. A real time shadow detection algorithm is presented to solve this problem. Firstly, the image is divided into image blocks based on inter-frame movement feature of sequence images in this algorithm. Secondly, the algorithm judges if this image frame is obscured and detects the possible shadowed area with the inter-frame predictor based on the gray mean of the image block. Finally, the shadow area is confirmed with a space domain filter for eliminating the erroneous detection areas. The experiment shows that this algorithm is real time and feasible.