POST-PROCESSING METHOD FOR VECTOR QUANTIZATION TO ACHIEVE A HIGHER PSNR AND NEARLY CONSTANT BIT RATE

摘要

In a conventional vector quantization (VQ) encoding method, the reachable best PSNR can be realized by full search (FS) equivalent method when codebook is fixed. To achieve even higher PSNR and almost not degraded bit rate, a post-processing method for VQ encoding is proposed in this paper.rnThis method first checks how well the matching is for an image block after conventional VQ by its block PSNR_b. Then it partitions a bad matched image block into 2 partial blocks by a certain pattern (a pair of masks) further and encodes them again just within a small already encoded nearby area. The overall matching can generally be made better due to using 2 smaller partial blocks. Meanwhile bit rate can be kept nearly constant by using 2 shorter winner pointers instead of 2 winner indexes finally. At the receiver, the bad matched image block is synthesized in reverse way to realize a better reconstruction.rnIn the case of low-detailed image Splash, image PSNR can be improved more than 1 dB for 1024 codebook while bit rate has almost not degraded.