Robust deep kernel-based fuzzy clustering with spatial information for image segmentation

Clustering algorithms with deep neural network has attracted wide attention to scholars. A fuzzy K-means clustering algorithm model on adaptive loss function and entropy regularization (DFKM) is proposed by combining automatic encoder algorithm. Although it introduces improve the robustness of model, its segmentation effect not ideal for high noise. The research purpose this paper focus anti-noise performance image segmentation. Therefore, basis DFKM, segmentation, combine neighborhood median mean information current pixel, introduce membership degree, extend Euclidean distance kernel space using function, propose a dual-neighborhood constrained based (KDFKMS). large number experimental results show that compared DFKM classical algorithms, stronger robustness.