DocumentCode
3421432
Title
Still image compression through exhaustive two-valued shape-adaptive searches
Author
Amorós, Maria Bras ; Gonzalez-Conjero, J. ; Guitart-Colom, Pere ; Serra-Sagristà, Joan ; García-Vílchez, Fernando
Author_Institution
Dept. of Inf. & Commun. Eng., Univ. Autonoma de Barcelona
fYear
2006
fDate
28-30 March 2006
Lastpage
441
Abstract
Summary form only given. We adapted shape-adaptive coding and the BISK algorithm to new sign and refinement encoders, with the novelty of encoding separately the refinement bits for each set of coefficients having the same prefix. This allows the algorithm to capitalize the refinement redundancy among each of these sets. The proposed sign and refinement encoders may be considered independently and may be integrated to other bit plane encoders. However, the search scheme suggests a new complete bit plane encoder defined by an exhaustive two-valued shape-adaptive search (ETSE). Even though ETSE does not include arithmetic coding, the coding performance of ETSE is competitive when compared to other wavelet-based encoders which include arithmetic coding. Furthermore, as BISK, ETSE may perform compression of images with non-regular boundary
Keywords
arithmetic codes; binary codes; data compression; image coding; wavelet transforms; arithmetic coding; binary set splitting with k-d trees; bit plane encoders; exhaustive two-valued shape-adaptive search; image compression; refinement encoders; wavelet-based encoders; Arithmetic; Bicycles; Data compression; Decoding; Geoscience and remote sensing; Image coding; Image processing; PSNR;
fLanguage
English
Publisher
ieee
Conference_Titel
Data Compression Conference, 2006. DCC 2006. Proceedings
Conference_Location
Snowbird, UT
ISSN
1068-0314
Print_ISBN
0-7695-2545-8
Type
conf
DOI
10.1109/DCC.2006.74
Filename
1607284
Link To Document