Fractal Image Coding Using Fractional Differencing Model

We propose a fractal image coding scheme, which uses the Fractional Differencing Model (FDM) for the estimation of the fractal dimension and the scale parameter related to the Fractional Brownian Motion Process (FBMP). FDM has advantage over FBMP. That is, FBMP is known to be difficult to estimate because of its non-stationarity, continuity, while FDM can be considered as the discrete version of FBMP, thus, share all properties of FBMP and has simple estimator based on least squares method. Therefore if this estimated fractal parameter which represents the scale property in FBMP is used for encoding and to restrict the searching area for block searching as a classification measure, the computational burden in encoding process can be dramatically reduced. In this paper, we propose a searching algorithm and a non-searching algorithm using the FDM and compare the existing algorithms with our proposed schemes.