Automatic Segmentation of Moving Objects in Video Sequences Based on Dynamic Background Construction
Wu Si,LIN Shou-Xun,ZHANG Yong-dong
DOI: https://doi.org/10.3321/j.issn:0254-4164.2005.08.018
2005-01-01
Chinese Journal of Computers
Abstract:In this paper, an automatic segmentation algorithm for moving objects in video sequences based on dynamic background construction is proposed. First, based on a coarse separation of foreground from images, the dynamic background construction technique, an accurate background for current frame with information in temporally adjacent frames is constructed. Then, the intact moving objects’ areas are obtained by subtracting the constructed background from the current frame and merging the static foreground areas (namely, parts of objects stopping moving between adjacent frames) detected using the temporal information. Finally, the accurate objects’ contours are extracted by active contour (snake) using color gradient as its external energy. Experiments demonstrate that the proposed method is robust to the influence of uncovered background and objects’ irregular motion, and able to extract moving objects successfully from video sequences with dynamic background.