A fast BMA based on combining search candidate subsampling and APDS

Author

Jiang, Wenbin ; Zhou, Manli

Author_Institution

Electron. & Inf. Eng. Dept., Huazhong Univ. of Sci. & Technol., Wuhan

Volume

2

fYear

2004

fDate

30-30 June 2004

Firstpage

1115

Abstract

A new faster block-matching algorithm (BMA) named SSC-APDS is presented by subsampling search candidates in adjustable partial distortion search (APDS). Firstly APDS is modified to visit about half points of all search candidates by taking subsampling on them, using a spiral-scanning path with one skip. Two selected candidates that have minimal and second minimal block distortion measures are obtained Then a fine-tune step is taken around them to find the best one, while at most 8 more candidates were required. Experimental results show that the SSC-APDS can maintain its MSE performance very close to that of the APDS with higher speedup ratio. Moreover, the wider the search window is, the better SSC-APDS performs

Keywords

image matching; image sampling; mean square error methods; motion estimation; search problems; video coding; APDS; BMA; MSE performance; SSC-APDS; adjustable partial distortion search; block distortion measures; block motion estimation; block-matching algorithm; search candidate subsampling; search window; spiral-scanning path; video coding; Accuracy; Avalanche photodiodes; Computational complexity; Degradation; Distortion measurement; Encoding; Mean square error methods; Motion estimation; Q factor; Video sequences;

fLanguage

English

Publisher

ieee

Conference_Titel

Multimedia and Expo, 2004. ICME '04. 2004 IEEE International Conference on

Conference_Location

Taipei

Print_ISBN

0-7803-8603-5

Type

conf

DOI

10.1109/ICME.2004.1394406

Filename

1394406