Hyperspectral image super-resolution optimization method based on deep closed-loop neural network

Abstract

The invention discloses a hyperspectral image super-resolution method based on a deep closed-loop neural network. The method comprises two processes, namely constructing a deep closed-loop neural network model for hyperspectral data, and reconstructing a high-resolution hyperspectral image through variable separation and fine optimization. Two deep learning models are constructed to learn a super-resolution process and an inverse super-resolution process respectively, and form a closed-loop network to reduce a mapping space, and promoting model fitting; a network structure suitable for a hyperspectral image is adopted to extract spatial features and spectral features, spatial information and spectral information are jointly reconstructed, and the quality of the image obtained through super-resolution is improved; and model iterative solution is performed by using the trained deep closed neural network and adopting a variable separation fine optimization method, and a reconstruction result is optimized. According to the method, the deep closed-loop neural network suitable for hyperspectral image super-resolution is used, the mapping space can be reduced through the closed-loop network, and a reconstruction result better than that of a one-way network can be obtained.

Description

technical field [0001] The invention belongs to the field of remote sensing image processing, and in particular relates to a hyperspectral image super-resolution optimization method based on a deep closed-loop neural network. Background technique [0002] With the development of remote sensing technology, satellite hyperspectral images are used for effective detection and object type identification. Hyperspectral images have broad application prospects in many aspects such as object identification and geological survey. However, limited by the spectral range and hardware imaging conditions, hyperspectral images often have low spatial resolution, and it is very difficult to accurately classify and identify hyperspectral images under the condition of low spatial resolution. Therefore, before classification and recognition, the spatial resolution of remote sensing images can be improved through software methods, that is, super-resolution reconstruction technology that obtains a...

AI Technical Summary

Problems solved by technology

However, there are still problems in the hyperspectral image super-resolution method based on deep learning: (1) When building a deep learning model, the mapping space from low-resolution hyperspectral images to high-resolution hyperspectral images is too large, and it is difficult to learn the correct (2) The amount of data in the data set is limited and cannot include all the ground objects. For the cases that are not included in the data set, the model cannot learn the correct mapping relationship, and the reconstructed high-resolution hyperspectral image Both spatial consistency and spectral consistency in

Images