Review of Reconstruction Methods for Spectral Snapshot Compressive Imaging

Snapshot compressive imaging (SCI) uses a 2D sensor to obtain higher dimensional data and then reconstructs the underlying high-dimension data by elaborate algorithms. Applying SCI to capture hyperspectral images is known as spectral SCI. Although this technique has been proposed for more than a decade, it has not been widely used, mainly because its reconstruction accuracy and reconstruction speed are not yet satisfactory, which is the research focus on spectral SCI. This paper investigates the literatures on reconstruction methods of spectral SCI, mainly involving coded aperture optimization, model-based reconstruction algorithms and deep learning-based reconstruction algorithms. In this paper, we also provide a summary of studies on noise modeling and denoising for reconstructed spectral SCI data.