Title :
A mesh-based method for wavelet video coding using edge-detection in low frequency subband
Author :
Miok Kim ; Nam Ling ; Ralston, J.D. ; Li Song
Author_Institution :
Dept. of Comput. Eng., Santa Clara Univ., Santa Clara, CA, USA
fDate :
Feb. 27 2013-March 1 2013
Abstract :
In this paper, we propose a mesh-based method for motion estimation and compensation over low frequency subband for wavelet video coding. We concentrate on edge-detection information and apply different parameters using edge magnitude and edge angle values to indicate curvature values for objects. This helps to detect motion patches between adjacent frames for meshes. We analyze and compare coding efficiency with a variety of video test sequences. Experimental results show that our mesh method improves coding performance for videos with less globally distributed motions compared to local motions.
Keywords :
edge detection; image sequences; mesh generation; motion compensation; motion estimation; video coding; wavelet transforms; edge angle values; edge magnitude; edge-detection information; low frequency subband; mesh-based method; motion compensation; motion estimation; motion patches; video test sequences; wavelet video coding; Discrete wavelet transforms; Image edge detection; Motion estimation; Video coding; Video sequences; Motion estimation; mesh-based motion estimation; video coding; visual communications; wavelet video coding;
Conference_Titel :
Circuits and Systems (LASCAS), 2013 IEEE Fourth Latin American Symposium on
Conference_Location :
Cusco
Print_ISBN :
978-1-4673-4897-3
DOI :
10.1109/LASCAS.2013.6519044