Encoded error calibration for a coded aperture spectrometer based on deconvolution

The principle of a 2D coded aperture spectrometer is described in this paper. The crosstalk of adjacent rows, which is caused by the optical system's point-spread function and the nonuniform illumination of the apertures, is the main source of the system decoded errors. Through the analysis of the effect of the crosstalk and nonuniform illumination on the decoded spectrum, the encoding matrix is modified. Based on the new encoding equation, an algorithm using Gold's deconvolution method is proposed to remove the crosstalk of adjacent rows. In the end, we evaluate the effect of this method through a series of contrast experiments.