Genetic algorithm using discrete cosine transformation for fractal image encode

Abstract

In this paper, a genetic algorithm using discrete cosine transformation is proposed to speedup the fractal encoder. By using discrete cosine coefficients, the optimal Dihedral transformation between the range block and domain block can be found to save a large number of the redundant MSE computations. Moreover, combining the discrete cosine transformation technique with the genetic algorithm, the length of the chromosome is shortened to smooth the landscape of the search space since the optimal Dihedral index was determined. Hence the encode velocity is accelerated further. Experiments show that the encoding speed of the proposed method is 100 times faster than that of the full search method, while the cost is the 1.1dB loss at the retrieved image quality.