当前位置: X-MOL 学术Sci. China Inf. Sci. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Learning hyperspectral images from RGB images via a coarse-to-fine CNN
Science China Information Sciences ( IF 7.3 ) Pub Date : 2021-09-06 , DOI: 10.1007/s11432-020-3102-9
Shaohui Mei 1 , Yunhao Geng 1 , Junhui Hou 2 , Qian Du 3
Affiliation  

Hyperspectral remote sensing is well-known for its extraordinary spectral distinguishability to discriminate different materials. However, the cost of hyperspectral image (HSI) acquisition is much higher compared to traditional RGB imaging. In addition, spatial and temporal resolutions are sacrificed to obtain very high spectral resolution owing to the limitations of sensor technologies. Therefore, in this paper, HSIs are reconstructed using easily acquired RGB images and a convolutional neural network (CNN). As a result, high spatial and temporal resolution RGB images can be inherited to HSIs. Specifically, a two-stage CNN, referred to as the spectral super-resolution network (SSR-Net), is designed to learn the transformation model between RGB images and HSIs from training data, including a band prediction network (BP-Net) to estimate hyperspectral bands from RGB images and a refinement network (RF-Net) to further reduce spectral distortion in the band prediction step. As a result, the learned joint features in the proposed SSR-Net can directly predict HSIs from their corresponding scenes in RGB images without prior knowledge. Experimental results obtained on several benchmark datasets demonstrate that the proposed SSR-Net outperforms several state-of-the-art methods by ensuring higher quality in HSI reconstruction, and significantly improves the performance of traditional RGB images in classification.



中文翻译:

通过从粗到细的 CNN 从 RGB 图像中学习高光谱图像

高光谱遥感以其非凡的光谱可区分性来区分不同材料而闻名。然而,与传统的 RGB 成像相比,高光谱图像 (HSI) 采集的成本要高得多。此外,由于传感器技术的限制,牺牲空间和时间分辨率来获得非常高的光谱分辨率。因此,在本文中,使用容易获取的 RGB 图像和卷积神经网络 (CNN) 重建 HSI。因此,高空间和时间分辨率的 RGB 图像可以继承到 HSI。具体来说,一个两阶段的 CNN,称为光谱超分辨率网络 (SSR-Net),旨在从训练数据中学习 RGB 图像和 HSI 之间的转换模型,包括一个波段预测网络 (BP-Net) 来估计来自 RGB 图像的高光谱波段和一个细化网络 (RF-Net) 以进一步减少波段预测步骤中的光谱失真。因此,所提出的 SSR-Net 中学习到的联合特征可以直接从 RGB 图像中的相应场景中预测 HSI,而无需先验知识。在几个基准数据集上获得的实验结果表明,所提出的 SSR-Net 通过确保更高的 HSI 重建质量优于几种最先进的方法,并显着提高了传统 RGB 图像在分类中的性能。所提出的 SSR-Net 中学习到的联合特征可以直接从 RGB 图像中的相应场景中预测 HSI,而无需先验知识。在几个基准数据集上获得的实验结果表明,所提出的 SSR-Net 通过确保更高的 HSI 重建质量优于几种最先进的方法,并显着提高了传统 RGB 图像在分类中的性能。所提出的 SSR-Net 中学习到的联合特征可以直接从 RGB 图像中的相应场景中预测 HSI,而无需先验知识。在几个基准数据集上获得的实验结果表明,所提出的 SSR-Net 通过确保更高的 HSI 重建质量优于几种最先进的方法,并显着提高了传统 RGB 图像在分类中的性能。

更新日期:2021-09-10
down
wechat
bug