Abstract: Although many mathematical and signal-processing tools exist, detection of sperm whales based on their sound recordings proves somewhat difficult. This paper presents some of the advantages of the wavelet transform and the spectrogram in analyzing sperm whale clicks. The coefficients of the wavelet transform and the short-time Fourier transform are used to provide a representation of the intrinsic characteristics of the sound emissions of the sperm whale. For detection, we propose a new parameter—Short-Time Windowed Energy—that characterizes the particular shape in the time-frequency/scale domain of sperm whale clicks. This paper illustrates the resistance to noise of this parameter. In addition, thanks to new processors, this algorithm, which was once lengthy in calculation time, can be integrated easily in a real-time system.
Key Words: SPERM WHALE; PHYSETER; CLICK; DETECTION; WAVELET TRANSFORM; FOURIER TRANSFORM
Document Type: Research article
Page Numbers: 463 - 467