彩色图像序列中运动人体轮廓提取

在视频序列的人体运动分析中,实时提取出运动人体轮廓,是很多研究起始的关键步骤.而彩色图像由于具有比灰度图像更多的视觉信息,受到了越来越多的重视.采用了一种新的色彩背景模型;运用改进的背景差分方法在复杂背景下获得运动人体的轮廓.实验结果表明上述算法对噪声抑制和人体图像断裂处填充都是有效的,在目标物是运动物体,且背景绝大多数均为静止时,该算法适用,能够实时提取出运动人体的轮廓.

Extract moving human silhouette from image sequences

Segmenting the motion human body in real-time is the first key step in the field of the analysis of human motion in the video sequence. Color images with more vision information than gray ones are becoming increasingly concerned. A new approach of constructing colorific background model is used, the moving human silhouettes is got in complex background using an improved background subtraction algorithm. The experimental results demonstrate that the algorithm is efficient for noise suppression and image gap filling. The algorithm is adapted when background is still in general and object is moving, it can extract the motion human silhouettes in real time.