The actual scientific task of developing methods of thematic segmentation of images from on-board systems of optical-electronic observation on the basis of ant algorithms is solved in the thesis. The object of research is the process of thematic segmentation of images from on-board systems of optical-electronic observation. The subject of research is the methods of thematic segmentation of images from on-board systems of optical-electronic observation. The aim of the work is to improve the quality of processing of images from on-board systems of optical-electronic observation by applying methods of thematic segmentation of images based on ant algorithms. Methods of probability theory, mathematical statistics, data clustering, evolutionary calculus, quantum methods, mathematical apparatus of the theory of matrices, methods of differential calculus, mathematics digital modeling, digital imaging, mathematical logic, information technology, optimization theory, methods for digital imaging, mathematical modeling, analytical and empirical methods of comparative research. Results and their novelty: the method of thematic segmentation of images from on-board systems of optical-electronic observation on the basis of an ant algorithm is improved, which, unlike the known ones, takes into account the peculiarities of image formation, the ant algorithm is used for thematic segmentation, and the segmentation of the image is reduced to the calculation of the target function, the totality the areas of motion of agents, the concentration of pheromone on the routes of agents, which increases the quality of image processing. The further development of the method of large-scale image processing from on-board systems of optical-electronic observation, in which thematic segmentation of images with different scale values is made on the basis of an ant algorithm, unlike in the known ones, and the scaling of segmented images with different scale values to the original size and filter image is calculated, and the resulting segmented image is a pixel-by-pixel product of the original image and filter image, which reduces errors and increases information indicators of thematic segmentation of images. The applied information technology for the thematic segmentation of images from on-board systems of optical-electronic observation is improved, which, unlike the known ones, are supposed to determine the quantity and magnitude of images, the color representation of images, the allocation of the brightness channels of each color space, the application for the thematic segmentation of image-based methods on the basis ant algorithms. The obtained practical results are confirmed by the acts of implementation and prove the correctness of the theoretical positions of the thesis, high efficiency of the developed methods and information technology.
