Iterational method for piecewise constant images restoration with an a priori knowledges of image objects boundaries

This article observes restoration method for piecewise constant images with predefined areas' borders based on Van Cittert filtration. As a distortion Gaussian and rectangle impulse response are considered. Image to be reconstructed is the result of sampling and noise addition. The results of experimental research of dependency of accuracy and stop-criterion from iteration number are demonstrated. Also potential application for remote sensing image analysis such as quality enhancement of average assessments is described.