The application of convolutional neural networks for tomographic reconstruction of hyperspectral images

• Published in: Displays Vol. 74; p. 102218
• Main Authors: Huang, Wei-Chih, Peters, Mads Svanborg, Ahlebæk, Mads Juul, Frandsen, Mads Toudal, Eriksen, René Lynge, Jørgensen, Bjarke
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2022
• Subjects: Computed tomography imaging spectrometer (CTIS) images; Convolutional neural networks; Hyperspectral cubes reconstruction; Convolutional neural networks; Hyperspectral cubes reconstruction; Computed tomography imaging spectrometer (CTIS) images
• ISSN: 0141-9382; 1872-7387
• DOI: 10.1016/j.displa.2022.102218

A novel method, utilizing convolutional neural networks (CNNs), is proposed to reconstruct hyperspectral cubes from computed tomography imaging spectrometer (CTIS) images. Current reconstruction algorithms are usually subject to long reconstruction times and mediocre precision in cases of a large number of spectral channels. The constructed CNNs deliver higher precision and shorter reconstruction time than a sparse expectation maximization algorithm. In addition, the network can handle two different types of real-world images at the same time—specifically ColorChecker and carrot spectral images are considered. This work paves the way toward real-time reconstruction of hyperspectral cubes from CTIS images. •A novel method, using CNNs, is proposed to reconstruct 3-D cubes from CTIS images.•The network can attain good accuracy for 5 and 25 spectral channels.•Reconstruction time is around 13 ms.•This work paves the way to real-time reconstruction of hyperspectral cubes.