The Key Technology Research Of2-D Slit-Array Hadamard Spectrometer

In order to balance the contradictions between the resolution and the signal-to-noise ratio of the grating dispersion spectrometer, our group brought up a new design of the mask of2-D slit-array Hadamard spectrometer with the help of MEMS technology and the research on the theory and key technique of the2-D slit array Hadamard spectrometer based on this new idea have been made in this paper.Having made analysis on how the2-D slit-array mask realize the Hadamard transform and how to get a signal-to-noise ratio gain without decrease the resolution of the spectrometer though Hadamard transform, the relationship between the order of the mask and the gain on the signal-to-noise ratio is brought up. After this, we discussed how to use the2-D slit-array Hadamard spectrometer so that we can achieve a reasonable gain on signal-to-noise ratio. Subsequently, the way to design the mask has been given in this paper, by which we can know how to set the width and length of the slit, the column distance and encoded matrix. A preliminary study on the relationship between the detection errors of the spectrometer and the machining errors of the mask has been made through the analysis on the influence of the machining errors of the mask on the measurement of the spectrometer. We set up a Hadamard spectrum-detecting system with a2-D slit-array mask, and successfully decoded the spectrum of Hg-lamp from the experimental data achieved from the spectrometer which has a high signal-to-noise ratio. We choose a matrix CCD as the detector of the spectrometer so that we can collect all the data at the same time. However, smearing, which is an inevitable phenomenon that occurs during the charge transfer process of CCD imaging sensors, has adversely affected the encoded and decoded process of the Hadamard spectrometer. We proposed two smear removal algorithms, one of which uses the optical black region(OBR), the other is based on the wavelet transform. The differences of the two algorithms are discussed through a lot of experiments.