Hyperspectral imaging (HSI) technology has become prominent, with a wide range of applications: food quality control,crop monitoring, and medical diagnostics. As HSI is able to capture spatial and spectral data, it is highly desirable, buthighly complex. However, this functionality presents a challenge for data acquisition as three-dimensional HSI imagesmust be acquired by an image sensor of one less dimension. Thus, HSI systems are often pushbroom systems, with twodimensionalimages being successively constructed over time from line scans. Additionally, HSI is expensive and difficultto operate. A snapshot HSI system is developed to address these challenges, whereby the additional image dimension isencoded onto an occupied dimension on the image sensor. Additionally, the snapshot HSI system is constructed from lowcost, readily available components. The presented snapshot HSI system consists of a transparent diffraction optical discbonded to an aperture mask, with alternating transparent and opaque regions, acting as an optical chopper when rotated bya DC brushless motor. This allows separation of the spectra of overlapped pixels on the HSI image sensor. When anincident beam passes through this optical chopper, many frequencies (corresponding to spatial channels) are imposed bythe binary mask, while undergoing diffraction across the visible spectrum. Overlapped spectra are directed at a chargecoupled device, where Fourier analyses distinguish each spatial channel. System geometry is used to transform the Fourieramplitude spectra into functions of wavelength for each spatial pixel. The design is experimentally validated throughcomparison to a commercially available spectrometer.
展开▼
