The study of the noise immunity of chip spectrometer based on sparse recovery
Feiying Sun, Zheng Yang, Peng Wu, Liangping Xia, Xiuhui Sun, Changbin Nie, Jun Cui, Shaoyun Yin, Chunlei Du · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2016
The noise immunity of spectral detection is analyzed theoretically for the kind of chip spectrometer based on the combine of integrated filter matrix and sparse recovery. With the sparse representation algorithm, the mathematic model of the detection noise and condition number of filter matrix is built for the Gaussian shaped incident spectrum. In the simulations, the filter matrix with different condition number is produced by changing the cavity length with the fabry perot (FP) interference model. An exponential relationship between the spectrum recovery variance and the condition number is obtained, which is consistent with the built theory model. The result reveals that when the condition number of filter matrix is 103 level, the noise tolerance of chip spectrometer will be as large as 10% when the variance is about 5 ×10-3 in the sparse recovery.