POWER DOMAIN ALGORITHMS FOR FRACTAL IMAGE DECOMPRESSION

摘要

A method of decompressing a greyscale or colour image stored in Iterated Function Systems (IFS) format with probabilities comprises applying each transformation in turn to the unit square to produce a series of parallelograms, each of which is contained within the unit square. Each transformation is then applied to each of the parallelograms, to produce a second level of parallelograms, all of which will be smaller than and contained within the parent parallelograms. The process is repeated until the parallelograms shrink to the size of a single pixel. The colour or intensity of each pixel of the decompressed image can then be calculated as a function of the number of parallelograms which have shrunk onto that particular pixel, weighted according to a parallelogram weighting factor. Black and white images can be accommodated by ignoring the probabilities.