Vorobel' R. Digital image processing based on the theory of contrast

Digital halftone images are the research object of the thesis. The aim of the research object of the thesis are: development of the theoretical base of images, which use theory of contrast as instrumentation tools for obtaining of perception independent quantity evaluation of the local contrast and images characteristics, construction on this base information technologies and hardware for images quality enhancement in the real time. Expression for computation of general contrast for subject images is constructed. Linear relation between local contrast and Laplasian is obtained. It is shown that the local contrast is a base for the construction of methods for images visual quality enhancement. General methods for global and moving stretch of the band of the brightness and histograms transformation are proposed. Euler’s transformation, beta-functions, linear and nonlinear piecewise functions for transformation of the components in this methods are used. General methods of moving weighted rank transformat ion are built. Novel classes of methods for unsharp masking using nonlinear transformation of background component and methods for the nonlinear transformation of the local contrast are obtained. Method for visual quality evaluation was . proposed. Technique of image processing and problem-oriented systems are implemented in industry.