Fully Automated Clustering based Blueprint for Image Analysis

Abstract

Data points are grouped together during clustering. The data points may be grouped according to comparable attributes using clustering methods. Data points are grouped using fuzzy clustering, which groups data points into one or even more clusters. Density Peak (DP) grouping may identify clusters, however as the sum of clusters is raised, memory overflow occurs because a normal-sized picture with more pixels is utilized for image segmentation, leading to a high level of similarity matrix. Automated Fuzzy Clustering Frame (AFCF) for picture segmentation might be used to prevent this. This framework offers three contributions. In order to lower the length of the similarity measure and hence increase the computational efficiency of the DP algorithm, the Density Peak approach is first employed for the idea of Super Pixel. A stable choice graph is produced by using the Density Balance approach, which also allows the DP algorithm to perform completely independent clustering. Last but not least, the system uses a Fuzzy c-means grouping based on previous entropy to enhance the results of picture segmentation. This allows for better segmentation outcomes by taking into account the data of pixels from spatial neighbors. The goal of the current study is to create and describe an Automated Fuzzy Clustering Framework for segmenting photos.