An improvement image vector quantization based on affine transformation

In this paper, the commonly used affine transformation (AT) in fractal coding technique is incorporated into image vector quantizer (VQ). Two new improved VQ coding schemes based on AT are investigated in detail without modifying the existing codebook. Experimental results show that the proposed method can obtain a better reproduced image as compared with the conventional VQ in terms of PSNR and visual quality even with a small size codebook. In addition, it does not need to generate an enlarged or specific codebook and has the advantage of low memory requirement