Accurate Spectral Super-resolution from Single RGB Image Using Multi-scale CNN

Published in Chinese Conference on Pattern Recognition and Computer Vision (PRCV), 2018

Paper (preprint)

Paper (camera ready)

PRCV 2018 poster

Code available on Github

Project webpabe

Abstract

Different from traditional hyperspectral super-resolution approaches that focus on improving the spatial resolution, spectral superresolution aims at producing a high-resolution hyperspectral image from the RGB observation with super-resolution in spectral domain. However, it is challenging to accurately reconstruct a high-dimensional continuous spectrum from three discrete intensity values at each pixel, since too much information is lost during the procedure where the latent hyperspectral image is downsampled (e.g., with ×10 scaling factor) in spectral domain to produce an RGB observation. To address this problem, we present a multi-scale deep convolutional neural network (CNN) to explicitly map the input RGB image into a hyperspectral image. Through symmetrically downsampling and upsampling the intermediate feature maps in a cascading paradigm, the local and non-local image information can be jointly encoded for spectral representation, ultimately improving the spectral reconstruction accuracy. Extensive experiments on a large hyperspectral dataset demonstrate the effectiveness of the proposed method.

Citation

@inproceedings{yan2018accurate,
title={Accurate Spectral Super-Resolution from Single RGB Image Using Multi-scale CNN},
author={Yan, Yiqi and Zhang, Lei and Li, Jun and Wei, Wei and Zhang, Yanning},
booktitle={Chinese Conference on Pattern Recognition and Computer Vision (PRCV)},
pages={206–217},
year={2018},
organization={Springer}
}