(2002) Image coding using entropyconstrained reflected residual vector quantization. Image Processing. 2002. Proceedings. 2002 International conference, 1.

PDF
14111_1.pdf Download (19kB)  Preview 

Microsoft Word
14111_2.doc Download (27kB) 
Abstract
Residual vector quantization (RVQ) is a structurally constrained vector quantization (VQ) paradigm. RVQ employs multipath search and has higher encoding cost as compared to sequential singlepath search. Reflected residual vector quantization (RefRVQ), a design with additional symmetry on the codebook, was developed later to a jointly optimized RVQ structure with singlepath search. The constrained RefRVQ codebook exhibits an increase in distortion. However, it was conjectured that the RefRVQ codebook has a lower output entropy than that of the multipath RVQ codebook. Therefore, the RefRVQ design was generalized to include noiseless entropy coding. We apply it to image coding. The method is referred to as entropyconstrained RefRVQ (ECRefRVQ). Since the RVQ scheme is able to implement very large dimensional vector quantization designs like 16/spl times/16 and 32/spl times/32 VQs, it is found highly successful in extracting linear and nonlinear correlation among image pixels. We intend to implement these large dimensional vectors with the ECRefRVQ scheme to realize a computationally less demanding imageRVQ design. Simulation results demonstrate that ECRefRVQ, while maintaining single path search, provides 1 dB improvement in PSNR for image data over the multipath ECRVQ.
Item Type:  Article 

Subjects:  Computer 
Department:  College of Computing and Mathematics > Mathematics 
Depositing User:  Mr. Admin Admin 
Date Deposited:  24 Jun 2008 13:23 
Last Modified:  01 Nov 2019 14:04 
URI:  http://eprints.kfupm.edu.sa/id/eprint/14111 