Filtration of images with constant brightness areas corrupted by additive noise

In this paper the algorithm of filtration of digital images that consisted of objects that described as areas with constant grey levels and corrupted by additive noise with uncorrelated samples is investigated. The developed algorithm uses the technique of two-stage processing allowing performing of digital images processing that decreases a computational complexity and in the same time preserve in some manner two-dimensional processing of an image. Results of simulation obtained while processing synthesized image. The comparison of the results of filtration made for different noise level. Comparison with known and widespread algorithms of image filtration is carried out using MATLAB.