Reconstruction of satellite images based on fractal filters

Receiving of remote sensed data's signals in urban space information reception centers is usually difficult, because of complex electromagnetic situation in cities and insufficient EMC. Traditional methods for digital reconstruction of images use smoothing and autoregressive forecasting. In this case anomalous spikes on the image are lost. Alternative method based on Kolmogorov-Wiener's filters and fractal properties of satellite images is proposed in this article. Offered method allows to predict pixel values of satellite images with normalized RMS error of the order of 20-30%.