Сегментація медичних зображень
Вантажиться...
Дата
2020
Автори
Мостовий, В.В.
Горященко, С.Л.
Mostovyi, V.V.
Horiashchenko, S.L.
Назва журналу
Номер ISSN
Назва тому
Видавець
Хмельницький національний університет
Анотація
В роботі розглянуто та проаналізовано можливості застосування методів сегментації на основі ознак
зв’язаності для інших типів зображень, проведений аналітичний огляд і наведена класифікація відомих методів
сегментації, на підставі чого сформульовані вимоги щодо розробки структурних моделей для задач сегментації
мікроскопічних медичних зображень, обґрунтована актуальність використання ознаки зв’язаності щодо задач
сегментації і побудовані її математичні моделі.
Segmentation is an integral part of the digital image processing process. It is the division or division of the image into some parts that meet the specified characteristics and characterize these areas and the image as a whole. At the segmentation stage, issues are solved that complement the standard tasks of image processing, namely coding, restoration, quality improvement. The segmentation process is considered an integral part of the tasks of image recognition, classification and identification. That is why segmentation has found its wide application in such areas as microbiology, medicine, astronomy, military equipment and other areas of human life. Such research also helps psychologists and physiologists to study such processes as the perception of forms, learning and recognition of objects by living organisms and the human brain, and so on. Segmentation is widely used in the automation of microscopic examinations of various medical objects, which include the processing of images of cells of organisms and their components and hemocytological drugs. This process is an integral part of recognition and classification in medical diagnostics. Recently, work has begun on the complete automation of the process of segmentation of images of biological objects in order to increase the reliability of the diagnosis of various diseases. The information obtained as a result of segmentation is also used to identify the effects of various adverse factors and helps to predict the course of leukemia, lymphosarcoma, anemia and other diseases of the human body. The article considers and analyzes the possibility of applying segmentation methods based on signs of connectivity for other types of images, conducted an analytical review and classification of known segmentation methods, based on which the requirements for developing structural models for segmentation of microscopic medical images, substantiated the relevance of the feature connections on segmentation problems and its mathematical models are built.
Segmentation is an integral part of the digital image processing process. It is the division or division of the image into some parts that meet the specified characteristics and characterize these areas and the image as a whole. At the segmentation stage, issues are solved that complement the standard tasks of image processing, namely coding, restoration, quality improvement. The segmentation process is considered an integral part of the tasks of image recognition, classification and identification. That is why segmentation has found its wide application in such areas as microbiology, medicine, astronomy, military equipment and other areas of human life. Such research also helps psychologists and physiologists to study such processes as the perception of forms, learning and recognition of objects by living organisms and the human brain, and so on. Segmentation is widely used in the automation of microscopic examinations of various medical objects, which include the processing of images of cells of organisms and their components and hemocytological drugs. This process is an integral part of recognition and classification in medical diagnostics. Recently, work has begun on the complete automation of the process of segmentation of images of biological objects in order to increase the reliability of the diagnosis of various diseases. The information obtained as a result of segmentation is also used to identify the effects of various adverse factors and helps to predict the course of leukemia, lymphosarcoma, anemia and other diseases of the human body. The article considers and analyzes the possibility of applying segmentation methods based on signs of connectivity for other types of images, conducted an analytical review and classification of known segmentation methods, based on which the requirements for developing structural models for segmentation of microscopic medical images, substantiated the relevance of the feature connections on segmentation problems and its mathematical models are built.
Опис
Ключові слова
сегментація, математичні моделі сегментації, імітаційне моделювання, segmentation, mathematical models of segmentation, simulation modelling
Бібліографічний опис
Мостовий В. В. Сегментація медичних зображень / В. В. Мостовий, С. Л. Горященко // Вісник Хмельницького національного університету. Технічні науки. – 2020. – № 5. – С. 51-56.