COLOR IMAGE COMPRESSION AND LIMITED DISPLAY USING SELF-ORGANIZATION KOHONEN MAP

Limited color displays are popular nowadays, Several methods have been proposed to do color quantization for limited color displays, Self-or ganization of Kohonen feature map (SOFM) is a very useful tool for dat a clustering, By extracting a special butterfly-jumping sequence from an image, a neural network was fed in that sequence of image data for SOFM training, and a fairly good color table was present to represent that image. The peak signal-to-noise ratio of the decoded image is hig h (about 35 dB in average for 256 color image), and the perceptual qua lity is good as well, even with a small set of color table (for exampl e, 32 or 64 colors), Furthermore, the training process is fast, For en coding, we also propose an efficient algorithm based on the Kohonen ma p ordering property, By considering that the neighboring image pixels are closely related and by setting some acceptable thresholds, we can quickly get an encoded image, For further compressing on the color ind exed image with the limited color palette, we cut the indexed images i nto 4 x 4 blocks and send the block vectors into another SOFM neural n etwork for training, Under the two-dimensional (2-D) mesh neural struc ture, SOFM vector quantization on an indexed image could largely reduc e the color shift artifacts and avoid the requantization problem, Abou t 0.5 b per pixel of coded image can be easily obtained and has a fair ly good perceptual quality, More importantly, the decoded color indexe d images can be readily displayed, This will reduce the decoder comple xity greatly.