Stochastic generation of the codebook in vector quantization of images: compression systems

A major objective of image coding is to represent an image with as few bits as possible while preserving the level of quality and intelligibility required for the given application. Among the most widely used schemes, vector quantization has received considerably attention. The vector quantization scheme has proven to be very effective in speech and image coding. One of the most important steps in the whole process is the design of the codebook. The codebook is generally designed using the LBG algorithm, that is in essence a clustering algorithm which uses a large training set of empirical data that is statistically representative of the image to be quantized. The problem that we are addressing in the paper is the stochastic generation of the codebook