Onishchenko V. Method of structural and linguistic images recognition for semantic transformation and collection of the relevant information in computer systems and networks

Research object: process of transformation and collection of relevant images of objects in computer systems and networks. Research purpose: development and grounds of the structural and linguistic recognition of images method for semantic transformation and collection of the relevant information in computer systems and networks on the basis of semantic network of concepts application. Research methods and equipment: the information theory, the theory of mathematical logic, the theory of mathematical linguistics, the theory of sets, the theory of automatic machines and formal grammar, the theory of images recognition, probability theory and the mathematical statistics, experimental researches. Theoretical and practical results: improvement of existing computer systems and networks on the basis of development of the new hardware-software means that raise efficiency of collection and the analysis of the graphic information relevant to the retrieval request for account of structural and linguistic recognition of images method application. Novelty: for the first time the mathematical model of semantic transformation of a contour image in structure of concept is developed, based on construction of a vector of structural transformations of the initial image and allows eliminating influence of affine transformations and deformation distortions of a contour on recognition process. For the first time the method of structural and linguistic recognition of contour images of the three-dimensional objects is developed, based on construction structural and linguistic concepts allowing essentially expanding a range of the images accepted to consideration taking into account different directions of shooting and different corners of a declination of the camera from position in nadir. A method of collection of the graphic information in the distributed and telecommunication networks relevant to the retrieval request received the further development that has allowed reducing time of collection of relevant images. Implementation field: Usikov Institute of Radio Physics and Electronics of NASU, the State enterprise "Kharkiv Research Institute of Mechanical Engineering Technology". Application field: for the systems of different functions connected with formation, processing, storage and use of images.