摘要:
A method is provided for constructing a video frame in a video signal processing apparatus using multi-core processing, in which multiple macro blocks are generated from a video signal on a predetermined unit basis, and a video frame is constructed by combining the generated multiple macro blocks with header information. The video frame includes one correlation information field for recording information about correlations between the multiple macro blocks, and the one correlation information field is located in front of the multiple macro blocks constituting the video frame.
摘要:
Disclosed is a method for detecting a scene change in real time in order to control a moving-picture encoding data rate, the method including: dividing a current frame into a plurality of regions, and calculating a dissimilarity metric (DM) of each divided region; determining if the dissimilarity metric of each divided region is beyond a preset reference value; calculating the number of regions, the dissimilarity metric of which is beyond the preset value, in the current frame; and determining that a scene change occurs in the current frame, when the number of regions, the dissimilarity metric of which is beyond the reference preset value, is equal to or greater than a preset threshold value.
摘要:
A method for estimating a boundary between a background and a user object in a video, in which the user object is segmented from the background is provided. The initial object contour is estimated in a first video frame for object segmentation, and color probability density functions are estimated for the user object and the background segmented from the first video frame. An edge strip with a predetermined width is generated in a current frame based on an object contour of a previous frame, and a coherence probability density function of the edge strip is estimated. The user object is segmented in the current frame by energy minimization.