Abstract
Image communications is primarily constrained due to its large bandwidth requirements. Therefore, researchers worked on various compression algorithm to achieve low bit rate. It was stated that images and video sequences are highly-correlated sources and their correlation should be exploited in a given compression algorithm Differential pulse code modulation (DPCM) has emerged as a mean of exploiting the correlation among the image pixels. Later on, DPCM was improved upon by predictive vector quantization (PVQ). PVQ employs block by block prediction and results in satisfactory performance at low bit rates. However, its design is complicated and recently an asymptotic closed-loop (ACL) was proposed to stabilize the design. In this paper, we attempted to replace the VQ with a multistage VQ structure in a hope to further reduce the stress on the closed-loop design. The multistage VQ structure that we employed is commonly referred to as reflected residual vector quantization (RRVQ). RRVQ works by imposing an additional symmetry constraint on the multistage codebook design. RRVQ has been quite popular where large block-length vector quantizations is needed due to their very low codebook search capability. Our proposed design goal in replacing VQ with RRVQ in a PVQ design is our wish to use large block length like 16 x 16 or 32 x 32 size vectors to grab any linear/nonlinear correlation among the vector components. The way to incorporate RRVQ within PVQ structure has been proposed and simulation results are discussed.