Title :
Motion compensation for video codec based on disparity estimation
Author_Institution :
City for Sci. Res. & Technol. Applic., Inf. Res. Inst. IRI, Alexandria, Egypt
Abstract :
The segmentation methods applied in MPEG-4 and H.26× showed mosquito and blocking effect in low bit rate applications. We have proposed a technique mainly directed to solve the mosquito and blocking effect of video coding. The procedure will be referred as Disparity Estimation based Video Codec (DEVC). First the frames are processed by binarization for separating the moving parts from the still parts in the frame; it is analogous to separating an object from the background in image processing. Second multi-resolution images will be created from the motion parts of the frame. The objects will be represented as region segments in each resolution. A matching or correspondence is done between pair of selected frames to compensate for the frames between them. Finally the motion vectors will be created by applying the “Steepest Decent Method”. DEVC showed comparatively high quality video sequences compared to MPEG-4. In addition the mosquito effect, blocking effect and The Motion Estimation (ME) mismatches have been removed from video sequences at low bit rate.
Keywords :
data compression; image representation; image resolution; image segmentation; image sequences; motion compensation; motion estimation; video codecs; video coding; DEVC procedure; H.26× coding; ME mismatch; MPEG-4 coding; binarization processing; bit rate application; blocking effect; disparity estimation based video codec; frame compensation; image processing; mosquito effect; motion compensation; motion estimation; motion vector; moving part separation; multiresolution image; object representation; segmentation method; steepest decent method; video codec; video coding; video compression; video sequences; Image segmentation; Motion compensation; Motion estimation; Motion segmentation; Transform coding; Vectors; Video compression; Law Bit Rate; Motion Compensation; QoS; Video compression;
Conference_Titel :
Computers and Communications (ISCC), 2012 IEEE Symposium on
Conference_Location :
Cappadocia
Print_ISBN :
978-1-4673-2712-1
Electronic_ISBN :
1530-1346
DOI :
10.1109/ISCC.2012.6249355