Browse

The main scientific problem is the construction of methods for describing and analysing images that consider multilevel periodicity and the dependence of their fuzzy properties on the scale of observation. The vast majority of digital image processing algorithms are based on a precise algorithm for extracting useful information, which often does not allow vision systems to solve nontrivial problems. To date, a vast number of "fuzzy" algorithms (based on the use of the theory of fuzzy sets) have been developed, all...

The main tasks and provisions of multiscale image processing of nanostructures are formulated, the feature of which is the use of the concepts of heredity and variability of characters. Heredity refers to the process of preserving traits on a sequence of images of a nanostructure as the scale of observation changes. Under variability refers to the process of loss of signs or the emergence of new signs on a multiscale sequence of images. It is obvious that images of nanostructures on scales at which there is a shar...

The sequence of images received when studying a nanoobject (the induced cluster-type nanostructures) at different scales is offered to consider as result of mutual influence of factors of heredity and variability of signs on some of his initial characteristics. Features of heredity and variability of signs gives the chance of creation of hierarchical structures of formal descriptions are described, i.e. models of the nanoimages based on groups of the signs relating to various morphological and large-scale levels o...

The task of the work is to improve the quality of digital image processing in vision systems by developing new features based on the use of fractal theory in conjunction with fuzzy logic and the theory of fuzzy sets. At present, the systems of technical vision are actively used in many fields of science and technology. The most important part of any such system is the software, whose task is to process the information received. However, the overwhelming majority of digital image processing algorithms are based on ...