Features kept generative adversarial network data augmentation strategy for hyperspectral image classification

Mingyang Zhang, Zhaoyang Wang, Xiangyu Wang, Maoguo Gong, Yue Wu, Hao Li

Abstract: Highlights•A novel data augmentation strategy called feature-preserving generative adversarial network data augmentation (FPGANDA) is designed to alleviate the small-sample and sample imbalance problems in hyperspectral image classification.•A novel generative adversarial network and a new band selection and mixture strategy are developed for synthesizing mixed data possessing the main features of original data and the diverse features of generated data.•Experimental results show that the proposed method can get competitive performance.