TY - GEN
T1 - Aliasing error reduction based fast VBSME algorithm
AU - Liu, Qin
AU - Huang, Yiqing
AU - Goto, Satoshi
AU - Ikenaga, Takeshi
PY - 2008/9/25
Y1 - 2008/9/25
N2 - Mathematical analysis reveals that high frequency signal components are the main issues that make MRF algorithm essential. Moreover, the aliasing problem of subsampling algorithm also comes from high frequency signal components. So based on these mathematical investigations, two fast VBSME algorithms are proposed in this paper, namely roberts cross edge detector based subsampling method and motion vector based MRF early termination algorithm. Experiments show that strong correlation exists among the motion vectors of those blocks belonging to the same macroblock. Through exploiting this feature, a dynamically adjustment of the search ranges of integer motion estimation is proposed in this paper. Combing our proposed algorithms with UMHS almost saves 96%-98% Integer Motion Estimation (IME) time compared to the exhaustive search algorithm the induced coding quality loss is less than 0.8% bitrate increase or 0.04 db PSNR decline on average.
AB - Mathematical analysis reveals that high frequency signal components are the main issues that make MRF algorithm essential. Moreover, the aliasing problem of subsampling algorithm also comes from high frequency signal components. So based on these mathematical investigations, two fast VBSME algorithms are proposed in this paper, namely roberts cross edge detector based subsampling method and motion vector based MRF early termination algorithm. Experiments show that strong correlation exists among the motion vectors of those blocks belonging to the same macroblock. Through exploiting this feature, a dynamically adjustment of the search ranges of integer motion estimation is proposed in this paper. Combing our proposed algorithms with UMHS almost saves 96%-98% Integer Motion Estimation (IME) time compared to the exhaustive search algorithm the induced coding quality loss is less than 0.8% bitrate increase or 0.04 db PSNR decline on average.
UR - http://www.scopus.com/inward/record.url?scp=52149093678&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=52149093678&partnerID=8YFLogxK
U2 - 10.1109/CISP.2008.337
DO - 10.1109/CISP.2008.337
M3 - Conference contribution
AN - SCOPUS:52149093678
SN - 9780769531199
T3 - Proceedings - 1st International Congress on Image and Signal Processing, CISP 2008
SP - 85
EP - 89
BT - Proceedings - 1st International Congress on Image and Signal Processing, CISP 2008
T2 - 1st International Congress on Image and Signal Processing, CISP 2008
Y2 - 27 May 2008 through 30 May 2008
ER -