Passive CFAR detection based on continuous wavelet transform of sound signals of marine animal

摘要

Based on the characteristics of short pulse signals in the marine animals sound and the characteristics of the high temporal resolution of the wavelet transform, this paper proposed a new strategy of Constant False Alarm Rate(CFAR) pulse detection based on continuous wavelet transform(CWT). The CFAR pulse detection was carried out on each scale of the results of time frequency analysis of CWT. Then the pulse detection results of each scale were integrated to get the final detection results. The processing results of the simulation data and the actual data contained sperm whale sound signal show that the CFAR pulse detection based on CWT is superior to that based on the Short Time Fourier Transform(STFT).