Fast Fractal Image Encoding Based on Root Mean Square Error
Tao Xu · Journal of Chinese Computer Systems · 2007
In order to solve the problem of time consuming in the encoding process of the basic fractal image compression method,a fast algorithm based on root mean square error is proposed.Every child or parent block is classified into smooth or non-smooth block according to its root mean square error.The codebook is made up of all of the non-smooth parent blocks,and ordered by their roots mean square error.If a child block is smooth,its mean will be kept,or the nearest neighbor will be found in the sense of root mean square error in the ordered codebook.When searching for the best parent block in the vicinity of the nearest neighbor,the eight isometric transformations are acted on every parent block.At the same time,an error threshold is used to control the searching range around the nearest neighbor.The experimental results demonstrate that the proposed algorithm is much faster than the basic fractal algorithm.Moreover,both the quality of decoded image in the premise of close encoding time and the encoding speed in the premise of close PSNR(peak signal-to-noise ratio) are better than these in the algorithm based on local cross trace.