Self-adaptive manifold discriminant analysis for feature extraction from hyperspectral imagery
作者：
Highlights：
• We propose a self-adaption optimization-based manifold learning method.
• A two-stage projection matrix optimization model to optimize the projection matrix for FE.
• Maximal manifold margin criterion is designed to quantify the similarity among embedded features.
• Backward propagation strategy is introduced to minimize the loss value through iteration process.
• Experiments validate the superior HSI classification performance.
摘要
•We propose a self-adaption optimization-based manifold learning method.•A two-stage projection matrix optimization model to optimize the projection matrix for FE.•Maximal manifold margin criterion is designed to quantify the similarity among embedded features.•Backward propagation strategy is introduced to minimize the loss value through iteration process.•Experiments validate the superior HSI classification performance.
论文关键词：Hyperspectral remote sensing,Feature extraction,Self-adaptive optimization,Manifold margin,Discriminant features
论文评审过程：Received 31 October 2019, Revised 19 March 2020, Accepted 3 June 2020, Available online 12 June 2020, Version of Record 19 June 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107487
