Investigation of a compressive line sensing hyperspectral imaging sensor

Abstract

Passive hyperspectral imaging (HSI) sensors are essential in many space-borne surveillance missions because rich spectral information can improve the ability to analyze and classify oceanic and terrestrial parameters and objects/areas of interest. A significant technical challenge is that the amount of raw data acquired by these sensors will begin to exceed the data transmission bandwidths between the spacecraft and the ground station using classical approaches such as imaging onto a detector array. In this paper, the Compressive Line Sensing (CLS) imaging concept, originally developed for energy-efficient active laser imaging, is extended to the implementation of a hyperspectral imaging sensor. CLS HSI imaging is achieved using a digital micromirror device (DMD) spatial light modulator. A DMD generates a series of 2D binary sensing patterns from a codebook that can be used to encode cross-track spatial-spectral slices in a push-broom type imaging device. A high sensitivity single-element detector can then be used to acquire the target reflections from the DMD as the encoder output. The target image can be reconstructed using the encoder output and the encoding codebook. The proposed system architecture is presented. The initial simulation and experimental results comparing the proposed design with the state-of-the-art are discussed.