UV Raman chemical imaging using compressed sensing

Abstract

Different chemical (hyperspectral) imaging techniques have proven to be powerful tools to provide and illustrate insightful data within a broad range of research areas. The present communication includes proof-of-principle results of UV Raman hyperspectral imaging, achieved via compressed sensing measurements using coded apertures (CA) and a reconstruction algorithm. The simple and cheap CA set up, obtained by a 50% overall transmissive random binary mask (chromium on fused silica with 100 μm × 100 μm pixel size) positioned at the entrance plane of an imaging spectrograph, resulted in an overall high throughput for the UV region of interest. The mask was mounted on a translation stage, allowing reproducible switching to different CA, thus making possible for multi frame CA imaging. Results from a scene containing liquid droplets are shown as examples and, as expected, qualitative improvements in resolution and contrast could be observed in both the spatial and spectral domain as the number of CA frames was increased.