Image data compression-encryption using G-SCAN patterns

Abstract

This paper presents an efficient image compression-encryption methodology by using fractals. The methodology is based on the principles and ideas reflected by the family of fractal based languages (Scan) and is mainly motivated for the compression and encryption of 2-D digital images. Scan represents a family special purpose context free languages which access sequentially or process the n×n data of a 2-D array, by describing and generating a wide range of (n×n)! algorithms (space filling curves, or permutations) from a short set of simple ones. The compression part of the Scan method uses the algorithmic accessibility of a 2-D image and applies a region searching coding criterion for the compression of the image data based on colors or gray levels. Scan produces several compressed versions, where the best of them (the one with the smallest number of bits/image) could be used for compression. The encryption part of the Scan method is coming directly from the nature of the Scan methodology, since it uses one of the (n×n)! permutations to compress the image. The Scan methodology can be used either as lossy or lossless compression scheme.