Processing Methods in Frequency Domain for Bubble Laser Scattering Signals

Liping Su,Weijiang Zhao,Deming Ren,Yanchen Qu,Xiaoyong Hu
DOI: https://doi.org/10.1117/12.834911
2009-01-01
Abstract:The key to the research on the bubble-laser-scattering property is to improve the signal to noise ratio. The various noises in the bubble-laser-scattering signal, the background light, power frequency of 50Hz, radio frequency noise and inherent noise within the electronic system, are analyzed in detailed. Two frequency-domain methods used to process the bubble scattering signal, the Fourier transform and the power spectral estimation, are focused on. These methods could reduce the noises as well as isolate the DC component and eliminate the impact of the power frequency of 50Hz, thus the pulses of the bubble laser scattering could be detected. The physical meaning for the power spectral estimation is clear. The power spectrum curve obtained by Yule-Walker AR (regression model) method is smoother and its frequency resolution is higher than the other methods. The results show that the Yule-Walker AR spectrum estimation is the most reasonable and most effective method.
What problem does this paper attempt to address?