Сегментація нечітких так адгезивних клітин грунтована на мінімальному кістяковому дереві та виявленні увігнутих точок

Abstract

The task of medical image segmentation has a wide range of applications in the field of medical research, such as clinical diagnostics, pathological analysis, surgical planning, computer surgery, and so on. In recent years, the global incidence of cancer is increasing, and early detection of cancer is particularly important. Accurate segmentation is an important part of computerized image analysis of blood cells. And the image of blood cells has characteristics of contact of cells, frequent strong adhesion, various sizes of cells, fuzzy border of cells and so on. It is difficult to accurately segment them. In particular, it has become a hot and complex topic for learning how to extract the cell area and achieve good segmentation of cell adhesion on a complex background.