An HVS based adaptive quantization scheme for the compression of color images
Digital Signal Processing
Hi-index | 0.00 |
A perceptual image coder for the compression ofmonochrome images is presented here, in which thecoding structure is coupled with a vision model toproduce coded images with an improved visualquality at low bit-rates. The coder is an improvementon the current image compression standard,JPEG2000. The proposed coding structureincorporates the human vision model in all stages ofcompression, and gives very good results compared tothe existing wavelet based JPEG scheme. Though themathematical model used are not new, the simplestructure of it which can be easily incorporated intothe coding structure of the JPEG2000, incorporatesthe visual processing stages in a very systematicmanner in the same way as the visual neurons processthe signal. The results presented in this paper revealthat the proposed structure gives very goodperceptual quality compared to the existing schemeespecially as we go for lower bit rates.