A fractal de-noising algorithm based on least absolute deviation method

Hui Guo, Jie He

School of Information and Electronic Engineering, Wuzhou University, Wuzhou 543002, Guangxi, China

Against the shortcoming that the traditional method of fractal image compression coding has inferior decoding quality on the original image subject to salt-and-pepper noise interference, this paper raises a least absolute deviation (LAD) method to be applied in fractal image compression, which can replace the method of least square error in computing contrast and brightness adjustment value and solve the L1-norm recursive problem using weighted median. The experimental result indicates that the LAD method has a very good anti-noise effect on the outliers introduced by salt-and-pepper noise.