Abstract
Hyperspectral image (HSI) super-resolution (SR) is an important technique for improving the spatial resolution of HSI. Recently, a method based on sparse representation improved the performance of HSI SR significantly. However, the spectral dictionary was learned under a fixed size, empirically, without considering the training data. Moreover, most of the existing methods fail to explore the relationship among the sparse coefficients. To address these crucial issues, an effective method for HSI SR is proposed in this paper. First, a spectral dictionary is learned, which can adaptively estimate a suitable size according to the input HSI without any prior information. Then, the proposed method exploits the nonlocal correlation of the sparse coefficients. Double ℓ 1 regularized sparse representation is then introduced to achieve better reconstructions for HSI SR. Finally, a high spatial resolution HSI is generated by the obtained coefficients matrix and the learned adaptive size spectral dictionary. To evaluate the performance of the proposed method, we conduct experiments on two famous datasets. The experimental results demonstrate that it can outperform some relatively state-of-the-art methods in terms of the popular universal quality evaluation indexes.
Highlights
Hyperspectral sensors can capture images with many contiguous and very narrow spectral bands that span the visible, near-infrared, and mid-infrared portions of the spectrum [1,2]
We propose an adaptive dictionary learning and double 1 regularized sparse-representation model for hyperspectral image (HSI) SR
The real HSI of these two datasets were treated as ground-truth, and they were used to produce the simulated low spatial resolution (LR) HSI and high spatial resolution (HR) multispectral image (MSI)
Summary
Hyperspectral sensors can capture images with many contiguous and very narrow spectral bands that span the visible, near-infrared, and mid-infrared portions of the spectrum [1,2]. Hyperspectral image (HSI) can provide fine spectral feature differences, to distinguish various materials, which can be widely and successfully used for many applications, such as object classification [3,4], tracking [5], recognition [6], and remote sensing [7,8]. Due to various hardware limitations, real captured HSI usually has low spatial resolution (LR), which significantly limits its application. It is not effective to enhance spatial resolution by improving the imaging quality of the hyperspectral sensors, and a breakthrough in hardware will be difficult and costly. HSI super-resolution (SR) has been proposed to generate a high spatial resolution (HR) HSI by fusing a high spectral resolution image, such as HSI, with an image, such as panchromatic image [9,10,11,12,13,14,15,16] or a multispectral image (MSI) [17,18,19,20,21,22,23,24,25,26,27,28,29,30]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
