DocumentCode :
2932707
Title :
Amalgamation of Singular Value Decomposition to JPEG for Enhanced Performance
Author :
Surobhi, Nusrat Ahmed ; Amin, Md Ruhul
Author_Institution :
Rajshahi Univ. of Eng. & Technol., Rajshahi
fYear :
2007
fDate :
7-9 March 2007
Firstpage :
295
Lastpage :
298
Abstract :
The demand of digital information compression is increasing dramatically because of the dominance of multimedia technology and the limitations of the physical media for handling huge amount of information. Compression reduces the storage and transmission burdens of raw information by reducing the ubiquitous redundancy without losing its entropy significantly. The image manipulation that occupies a significant position in multimedia technology necessitated the development of joint photographic experts group (JPEG) compression technique, which has proved its usefulness so far. Until recently, to minimize the blocking artifact, inherently present in JPEG at higher compression ratios, JPEG2000 is devised that makes use of wavelet function. In this work, a new approach to JPEG compression technique is proposed that enhanced the compression performances in comparison with aforesaid JPEG techniques. The new technique considers both discrete cosine transform (DCT) and singular value decomposition (SVD) method in the transform and reconstruction sides instead of using DCT only. The incorporation of SVD with nearest neighborhood approach has improved the compression performances significantly. A rigorous comparison of the various compression indices are made to validate the proposed algorithm. This approach is named as ´Hybrid JPEG´ (HJPEG) in this paper. The benchmark still image ´LENA´ is used for performance comparison.
Keywords :
data compression; discrete cosine transforms; image coding; singular value decomposition; JPEG; Joint Photographic Experts Group compression technique; discrete cosine transform; image manipulation; singular value decomposition amalgamation; ubiquitous redundancy; Communications technology; Discrete cosine transforms; Discrete transforms; Electronic mail; Equations; ISO standards; Image coding; Multimedia systems; Singular value decomposition; Transform coding;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information and Communication Technology, 2007. ICICT '07. International Conference on
Conference_Location :
Dhaka
Print_ISBN :
984-32-3394-8
Type :
conf
DOI :
10.1109/ICICT.2007.375396
Filename :
4261419
Link To Document :
بازگشت