Abstract: Aiming at detecting chromatic moving object from video sequences with fixed scene．HSV color space was adopted to represent each raw frame in video sequences，then it established background frame based on self-adaptive background model by extracting V value in raw frame, and detected moving object based on background subtraction．Motion region binarization incorporated H value in raw frame to filter motion region，in order to obtain the foreground frame in which the special moving object was selected．The experimental results show that the algorithm can achieve the special color moving object detection in the video with fixed scene and several moving objects very well．

Key words: Color space , Background-difference ,  Moving object detection

摘要： 针对固定场景视频序列中的彩色运动目标检测,采用HSV颜色空间表示视频原始帧,使用自适应背
景更新在原始帧中提取V 分量建立背景帧,利用背景差分法对视频序列进行运动检测,在运动区域二值化时结合原
始帧中的H 分量对运动区域进行筛选,从而得到特定运动目标被点亮的前景帧。实验结果表明,该算法能够有效地
实现在静止背景、多运动目标共存的视频中检测出特定颜色的运动目标。

关键词: 颜色空间 , 背景差分法 , 运动目标检测