Рақамли технологияларнинг назарий ва амалий масалалари 8-том 2-нөмір (2025) · 146-153-беттер
Parametric Optimization of Identification in Pollen Grain Image Recognition
Жуманов, И.И., Холмонов, С.М., Джуманов, О.И.
Аңдатпа
Methods for optimizing the identification of pollen grains using statistical, dynamic, textural, and image-specific features have been developed. Point-wise and nonlinear contour matching techniques for the input and reference objects-images of pollen grains-have been studied, along with the adjustment of raster parameters. Mechanisms have been implemented for reducing zero contour points, decreasing raster dimensionality, scaling, threshold and level control, encoding and placement of microobject images based on a pyramidal model, selection of reference contour points, cognitive analysis, and point search through annealing and prohibition, using stochastic modeling based on a truncated Markov chain. A computer-based identification system has been implemented in C++ using the CUDA parallel computing environment for the recognition, classification, and systematization of pollen grains.
идентификацияизображениепыльцевые зернараспознаваниеклассификацияпогрешностьтрудоемкостьстоимостькомплекс программidentification
Метадеректер дереккөзі: журналдың OAI-PMH архиві · Sindex толық мәтінді сақтамайды, дереккөзге сілтеме береді.