Рақамли технологияларнинг назарий ва амалий масалалари Том 8 № 3 (2025) · с. 29-37
Identification, Recognition, and Classification of Micro-Objects Based on Image Point Sparsification
Жуманов, И.И., Сафаров, Р.А., Джуманов, О.И.
Аннотация
Scientific and methodological principles for optimizing the processes of identification, recognition, and classification of micro-objects have been developed based on the use of component characteristics in the image structure. Tools for extracting statistical, dynamic, and morphological characteristics and a method for rarefying excess points on the surface of micro-objects have been proposed. Modified network training algorithms have been developed with tools for adjusting variable values, error control at the boundaries of acceptable values, and taking into account stationary, quasi-stationary, and non-stationary behavior of image points when forming training sets. The efficiency of the algorithms was studied according to the criteria of mean square error and information processing speed. A software package for visualization, recognition, and classification of pollen grain images was developed, the implementations of which were tested under conditions of a priori insufficiency, uncertainty, and non-stationarity of processes
изображениемикрообъектидентификацияраспознаваниеклассификациягибридная модельнейронная сетьоптимизацияпрограммный комплексimage
Источник метаданных: OAI-PMH архив журнала · Sindex не хранит полный текст, а даёт ссылку на источник.