An efficient bath fractal transform-based image coding technique

A computationally efficient image compression technique is presented, which is based on visual patterns as well as least squares fractal approximation by a self-affine system (SAS). Simulation studies demonstrate that the use of visual pattern-based technique can code the shade (flat) regions of an image more efficiently than fractal techniques. Furthermore, the proposed technique not only gives an encoding speed several times faster than other SAS-based fractal coding techniques but also achieves higher compression ratios and better fidelity.