Spatial and class structure regularized sparse representation graph for semi-supervised hyperspectral image classification

作者：

Highlights：

• Sparse representation based edge weighting method is employed in the graph based SSL.

• Spatial neighborhood information and probabilistic class structure are both incorporated into the sparse representation model.

• The proposed graph construction method is superior to state of the arts methods.

摘要

•Sparse representation based edge weighting method is employed in the graph based SSL.•Spatial neighborhood information and probabilistic class structure are both incorporated into the sparse representation model.•The proposed graph construction method is superior to state of the arts methods.

论文关键词：Spatial regularization,Probabilistic class structure,Sparse representation (SR),Semi-supervised learning (SSL),Hyperspectral image (HSI) classification

论文评审过程：Received 25 April 2017, Revised 17 January 2018, Accepted 27 March 2018, Available online 27 March 2018, Version of Record 6 April 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.03.027