A fractal block coding technique employing frequency sensitive competitive learning

The authors discuss a block oriented fractal coding technique for still images based on contractive affine transformation theory. A brief overview of the generalized technique is provided and a number of its attractive as well as less favorable features are discussed. In particular, the high order of computational complexity associated with the technique is addressed. A neural network paradigm known as frequency sensitive competitive learning (FSCL) is employed to assist the encoder in locating fractal self-similarity within a source image. A judicious development of the proper neural network size for optimal time performance is provided. Such an optimally chosen network has the effect of reducing the time complexity of the original encoding algorithm from O(n⁴) to O(n³). An implementation of fractal block coding employing FSCL and coding results is presented.