[23a-C302-1~9] 3.7 Optical measurement, instrumentation, and sensor (formerly 3.8)

△ [23a-C302-1] High-speed Fourier-transform spectral imaging with compressed sensing

〇(D)Shigekazu Takizawa¹, Kotaro Hiramatsu¹, Shunsuke Ono², Keisuke Goda^1,3,4 (1.Tokyo Univ., 2.Tokyo Inst. of Technology, 3.UCLA, 4.Wuhan Univ.)

Keywords:compressed sensing, Fourier-transform spectroscopy, Raman spectroscopy

Fourier-transform spectral (FTS) imaging is widely used for various range of scientific and industrial applications because it enables us to collect spatial and spectral information of the target sample. However, its spectral image acquisition rate is limited by the mechanical scanning speed of optical delay and further improvement of the measurement speed is required to measure more dynamic samples. In this report, we combined FTS imaging and compressed sensing to improve the spectral image acquisition rate. Specifically, we reconstructed spectral images from time-domain signals coarsely sampled by Lissajous scanning via sparse reconstruction. Our numerical simulations show that we can improve the spectral image acquisition rate by a factor of ten compared to a usual raster scanning. We also discuss robustness of the restoration results to noise and achievable compression ratios in this report.