Abstract: A novel video moving object segmentation algorithm based on spatio-temporal information was proposed in this paper. The algorithm can extract the moving object from the video sequence with static or global motion background automatically. Firstly, an efficient and accurate global motion compensation method was used to change the motion background to static background. In temporal motion information extracting,the value of background noise variance was estimated by histogram fitting to overcome the shortcoming of setting the value by experience, then the significance test and the symmetrical difference method were applied to achieve accurate moving o均ect mask. In spatial image information extracting, an improved multi-scale watershed algorithm based on viscous morphological gradient correction and edge value merging was employed to segment moving regions which can solve over segment problems greatly. Finally,video object was extracted by performing double threshold ratio operation on spatial and temporal results. Experimental results validate the proposed algorithm.

Key words: Moving object, Spatio-temporal Information, Motion compensation, Watershed algorithm