Publications Details
Real-Time Digital Signal Processing for a Fourier Transform Hyperspectral Imager
A Fourier Transform hyperspectral imager uses optical intereferometry to obtain hyperspectral data. Taking a Fourier Transform of the interferogram yields the frequency spectrum of the incident light. An optical system using a standard frame rate camera can generate such interferograms at a rate of 30 frames per second. Rather than store all of the raw interferogram data and process it afterwards, it is useful to have the ability to process the raw data in real time, generating and storing the hyperspectral data itself rather than the original interferograms. This real-time processing would result in a significant reduction in the data bandwidth and storage requirements, which are of particular interest in typical airborne environments with limited computing resources on board. This report details the digital signal processing algorithm and code developed for a processing subsystem based on the Texas Instruments TMS320C6201 fixed point processor. The function of this subsystem is to compute the magnitude Fourier Transform of the interferogram data at a rate of 30 frames per second.