A fast fractal decoding algorithm based on the selection of an initial image

In this correspondence, we show that the number of iterations required for the convergence of the fractal image decoding algorithm can be reduced by selecting a suitable initial image. We consider fractal decoding to have two components, namely, "DC" and "AC". Our initial image is an approximation of the attractor of DC decoding, which is a good estimation of the "range-averaged" image, i.e., the image obtained by replacing all pixel intensities in a given block by the mean value of the block. From the simulation, it is demonstrated that the decoding speed is greatly improved.