Efficient Image Segmentation Method Based on Sparse Subspace Clustering

摘要

A novel image segmentation method based on weighted sparse subspace clustering is presented. By choosing the l_(2,1) norm as sparse metric, feature datas are kept uniformly within the same subspace; By constraints of weighted sparse metric, feature datas are kept sparse within different subspace. Experiments show that the proposed weighted sparse subspace clustering method can obtain higher clustering accuracy than the state of old methods.