Komenchuk O. Intelligent information technology for object recognition in images using pseudomasks

Intelligent information technology for object recognition in images using pseudomasks The aim of the research is to increase the accuracy and speed of object recognition in images of small-sized, noisy and imperfect datasets by developing intelligent information technology using pseudomasks. The objectives of the research are: - to review known approaches and models for identifying and classifying objects in images of imperfect and small-sized datasets and to analyze the main problems that arise during the identification and classification of objects in images of small-sized and imperfect datasets; - to develop intelligent information technology, including new approaches, methods and algorithms that allow to increase the accuracy of identification and classification of given objects using pseudomasks in images of imperfect and small-sized datasets; - to create algorithmic and software for the application of the proposed intelligent information technology and methods for identifying and classifying given objects in images of imperfect and small-sized datasets; - develop Jupyter notebooks in Python for the application of intelligent information technology using pseudomasks for the identification and classification of objects in images of imperfect and small-sized data sets; - test and implement the developed intelligent information technology on real data and tasks. The object of the research is the identification and classification of objects in images of imperfect and small-sized data sets based on pseudomasks. The subject of the research is methods, mathematical and intellectual models of object recognition in images of imperfect and small-sized data sets based on pseudomasks. Research methods. Analysis of literary sources, open data, experimental research, system analysis, machine and deep learning methods, etc. A total of 9 scientific papers have been published on the topic of the dissertation research. 4 articles have been published in professional journals of Ukraine in technical sciences, and 5 abstracts of reports at scientific and practical conferences.