Hyperspectral image classification based on multiscale convolutional network

Abstract

Hyperspectral Image (HSI) classification is an important task in the field of Hyperspectral Image processing. However, the existing classification methods unable to solve the problems caused by hyperspectral image information redundancy, insufficient image feature utilization and Hughes phenomenon. Aiming at these three problems, a hyperspectral image classification algorithm based on deep learning is proposed. The Multiscale Convolutional Neural Network (MCNN) was used to excavate deep features and realize the learning of multiscale features. Then, the features of different scales were fused and classified. The results show that the proposed algorithm has higher classification accuracy than the traditional ones. Also, it has strong generalization ability and robustness. The effectiveness and feasibility of the proposed algorithm are fully verified.