Study On New Method Of Signal Detection Of CASS

The Controlled Accurately Seismic Source(CASS) is a new source of high repeatability, hasno harm on the surrounding medium and can be operated continuously for long time. It can beused to surveying crustal structures and monitoring the temporal variation of crustal mediacharacteristics. The vibrant signal radiated by CASS and recorded by instruments at seismicstations is continuous and Linearly Frequency Modulated(LFM). Because of the energy radiatedby it is quite small and send out by long time operation, the original record can’t be used directlywith the low signal to noise ratio(SNR) of long distance seismic stations records. So We need toconvert it to a sufficiently high SNR and pulse source record.Now, we commonly used matched filtering method to realize the signal detection and wavetransform. In consideration of the radiated signal is LFM, we try to use the combined ofWigner-Ville distribution and Hough transform method to achieve this goal.In this paper, we firstly introduce the history of linear time-frequency analysis from Fouriertransform and their merits and demerits. System introduced the bilinear time-frequencydistribution of Wigner-Ville distribution from definition properties discrete method advantage anddisadvantage. Independently introduced several time-frequency analysis methods and somedistribution belonging to the Cohen class, the spectrogram, pseudo Wigner-Ville distribution,Choi-Williams distribution and other kernel distributions. Here, we introduced the Houghtransform theory, the definition of Wigner-Hough transform. The corresponding computerprogram is worked out base on Matlab and C language. And applied to practical data processing toresult in good effect.By analyzing the simulated signal and practical data processing, we reached thatWigner-Ville distribution can analysis the non-stationary and time-varying signals in the timedomain and frequency domain. The Wigner-Ville distribution of LFM is a line in time-frequencyplane. Hough transformation is an image signal processing technology used to identify thegeometric feature. It detective the straight line or curve in image space by mapping from imagespace to the parameter space. So we can use the Wigner-Hough transform to realize the sourcesignal detective and wave transform. In actual signal processing, in order to solve the problem of calculation and storage inWigner-Ville distribution and Hough transform and improve the computation speed and veracityin the LFM signal parameter estimation, we segment the signal according the frequencymodulation rules and down sampling. While using Hough transform to detect straight line, we canalso fix the angle in actual operation.