Analysis the processing algorithm for the frequency measurement variance of the acousto-optic spectrum analyzer

When the acousto-optic device worked on the Bragg model, the non-liner affected the diffraction beam. There were some errors between the diffraction beam deflection peak position and the input signal's frequency, which reduced the frequency measure accuracy of the acousto-optic spectrum analyzer. Under the existing optical experimental platform, we eliminated the CCD background noise by reducing the threshold firstly, and then we processed the data by four methods, the peak value method, the Gaussian fitting method, the squared cancroids method and the Hilbert transform method. The least frequency measure variance is 31.8 KHz(2), the data processed by the Gaussian fitting method. It provides theoretical support for reducing the frequency measurement variance of acousto-optic spectrum analyzer.