Nitsyn D. Recognition fractal objects on low contrasting images

Dissertation is devoted to the questions of classification of radiograms which diagnostics of tuberculosis is carried out, on the basis of their fractal dimensional and application of Bayesian rule in multidimensional space of diagnostic signs. Theoretical bases of application of fractal geometry are in-process developed to the decision of tasks of recognition of patterns. The mathematical and algorithmic providing of the computer system of classification of radiograms, intended for diagnostics of tuberculosis is developed. Experimental verification of efficiency of the computer system of classification of radiograms is executed.