DocumentCode :
1739622
Title :
Fast binary pyramid motion estimation
Author :
Song, Xudong ; Chiang, Tihao ; Lee, Xiaobing ; Zhang, Ya-Qin
Author_Institution :
IVAST, Santa Clara, CA, USA
Volume :
2
fYear :
2000
fDate :
2000
Firstpage :
1127
Abstract :
A fast binary pyramid motion estimation (FBPME) algorithm is presented. The proposed FBPME scheme is based on the representation of set of binary features in terms of multiresolution layers, XOR Boolean block matching and an N-scale tiling search scheme. Each video frame is converted into a pyramid structure of K-1 binary layers with resolution decimation plus one integer layer at the lowest resolution. At the lowest resolution layer, the N-scale tiling search is performed to select the initial motion vector candidates. Motion vector fields are gradually refined with the XOR Boolean block matching criterion and the N-scale tiling search schemes in higher binary layers. FBPME performs several thousands times faster than the conventional full search block matching scheme at the same PSNR performance and visual quality. It also dramatically reduces the bus bandwidth and on-chip memory requirement. Moreover, the hardware complexity is low due to its binary nature
Keywords :
Boolean functions; bandwidth compression; computational complexity; data compression; image matching; image representation; image resolution; motion estimation; search problems; video coding; FBPME algorithm; N-scale tiling search; PSNR performance; XOR Boolean block matching; binary features representation; binary layers; bus bandwidth reduction; fast binary pyramid motion estimation; hardware complexity; image resolution; motion vector fields; multiresolution layers; on-chip memory requirement; resolution decimation; video frame; visual quality; Bandwidth; Filtering algorithms; HDTV; Motion estimation; PSNR; Partial response channels; Partitioning algorithms; Robustness; TV; Video compression;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal Processing Proceedings, 2000. WCCC-ICSP 2000. 5th International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-5747-7
Type :
conf
DOI :
10.1109/ICOSP.2000.891738
Filename :
891738
Link To Document :
بازگشت