Застосування кластеризації в задачах нормалізації зображень

Abstract

The development of image processing techniques leads to the emergence of all new approaches to solving segmentation problems. For example, the problem of separation of adjoining objects in the image is still relevant. Among data mining tasks, clustering tasks occupy a special place. It should be noted that when solving clustering problems, the most popular algorithms are based on optimal partitioning of a set of data into clusters. Such algorithms are aimed at grouping data into clusters so that the objective function of the partitioning algorithm reaches an extremum (minimum). In this regard, it is possible to distinguish the C-means algorithm.