An image segmentation method based on a modified local-information weighted intuitionistic Fuzzy C-means clustering and Gold-panning Algorithm

Abstract

The image segmentation method based on clustering analysis has the advantages of small sample space constraints and strong universality. As an unsupervised clustering algorithm, the fuzzy C-means clustering algorithm is widely used in practical engineering. However, it is still some shortcomings: the fuzzy C-means clustering algorithm is difficult to interpret the noise effectively, which makes it more sensitive to the noise, and the selection of key parameters has to be made by trial and error experiments, reducing the adaptability of the algorithm. Besides, its iteration process is heavily influenced by the initial clustering centers and easy to fall into local optimum. Therefore, an intuitionistic Fuzzy C-means clustering method, based on local-information weight, is proposed in this paper. By introducing the local-information weight, the proposed algorithm adjusts the local-information influence weight adaptively in fuzzy partition, which enhances its robustness to noisy images. Furthermore, a novel swarm intelligence algorithm, called the Gold-Panning Algorithm, is proposed to optimize the initial clustering centers and key parameters in the clustering algorithm. By utilizing the Gold-Panning Algorithm, the adaptability of the proposed clustering algorithm is further improved. In this paper, the proposed methods are explained in detail and compared with the existing methods to demonstrate its superior performance.