Method for segmenting motion object of compressed motion pictures
暂无分享,去创建一个
The present invention takes a lot of time in the process relates to a moving object segmentation method of compressed video, the method for dividing a conventional object after restoring the whole image used in the composite of the motion information and color information, and therefore to extract not only is there is a problem that requires a lot of computational complexity due to the processing for the entire image. The present invention considering the above problems by the steps of decoding a certain distance from the frame of the compressed video, and extracts the motion vector through a pre-compensation process the motion vector due to camera movement and; Further comprising: a size of the motion vector specifying a predetermined size or more area to the motion block and extracts the outline block diagram of the motion block; Extracting the edge in consideration of the directions of the outside blocks, and determines the contour of the object by using the extracted edge and; Consists of dividing the final motion object decodes the interior of the determined final contour has the effect of reducing the processing time by decoding only a part of the motion block for determining a motion objects as a motion vector of a certain interval, the frame and determining a contour.