基于分级的MRF视频运动检测算法及其实现

基于分级的MRF视频运动检测算法及其实现 Research and Implementation on Multiresolution Algorithm Based on MRF for Video Motion Detection 乔予杰 QIAO Yu jie 1 2 first-author 胡俊 HU Jun 1 2 苏祥芳 SU Xiang fang 1 2 孙洪 SUN Hong 1 2 茹国宝 RU Guo\|bao 1 2 武汉大学电子信息学院 武汉大学电子信息学院 School of Electronic Information,Wuhan University School of Electronic Information,Wuhan University 武汉大学电子信息学院,湖北,武汉,430072 武汉大学电子信息学院,湖北,武汉,430072 在阐述马尔可夫随机场 （MRF）视频运动检测算法理论的基础上 ,采用分级的方法 ,形成图像序列的时空多分辨率结构 ,大大优化了单一分辨率 MRF算法的初始化过程 ,并通过软件进行了实现 ,有效改进了单一分辨率算法检测的结果 .该算法应用于数字视频监控系统 ,可明显提高检测精度 ,具有较好的应用前景 Intrduce video motion detection algorithm based on Markov Random Field（MRF）. The MRF based motion detection algorithm has more robustness than the others, but it has its defaults yet. Multiresolution is used in the MRF based algorithm .The results show that it can improve the initialization of the algorithm and decrease the mistakes of video motion detection.So,the multiresolution algorithm based on MRF is bat around.Its realization in software is set forth.The multiresolution algorithm based on MRF has wide foreground of application in many fields. 图像序列 运动检测 分级 马尔可夫随机场 image processing motion dection multiresolution Markov random field（MRF） TN919.8 国家自然科学基金资助 （ 6 0 0 72 0 41） 2002-01-01 2021-04-01 1