Survey of image segmentation methods based on clustering
Guojian Cheng, Lianhong Liu · 2020
With the rapid development and application of clustering algorithms, the problem of image segmentation has attracted more and more scholars and industry attention. This paper studies the clustering-based image segmentation method, summarizes the basic idea, and divides it into two categories, namely partition method and agglomeration method. The typical algorithms Kmeans algorithm, CLARANS algorithm, mean shift algorithm and DBSCAN algorithm of these two methods are introduced in detail. The advantages and limitations of these four methods are analyzed. Finally, it points out the future research trend of clustering algorithms and gives conclusions.