Encoded Diffractive Optics for Hyperspectral Imaging

Recent hyperspectral imaging has been taking advantage of the development of diffractive optics elements (DOEs) to encode the spectral information in a single or multiple snapshots. Including these diffractive elements, together with previous codification technologies, has allowed the development of compact optical systems at the expense of computational algorithms for recovering spectral information from encoded optical measurements. Diffractive optics design methodologies improve the reconstruction quality by designing point spread functions with reduced dispersion effects. Even more, diffractive optics designs that achieve state-of-the-art performance have been proposed by leveraging available datasets and algorithms based on deep neural networks. This chapter describes the basics for modeling diffractive optical imaging systems and presents compact diffractive optical systems for hyperspectral imaging, the main diffractive optics design methodologies, and the computational algorithms employed to recover the encoded measurements.