Acoustic Signal Characterisation using multi-resolution transforms.
Most acoustic signals received by underwater systems in shallow waters are non-stationary and corrupted by unpredictable noise sources. In most cases, the noise has a dramatic influence on the performances of these systems. While classical methods often fail to characterise these noises in such an e...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Text |
| Language: | English |
| Subjects: | |
| Online Access: | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.225.1431 http://www.arl.nus.edu.sg/twiki/pub/ARL/BibEntries/Seekings2003b.pdf |
| Summary: | Most acoustic signals received by underwater systems in shallow waters are non-stationary and corrupted by unpredictable noise sources. In most cases, the noise has a dramatic influence on the performances of these systems. While classical methods often fail to characterise these noises in such an environment, recent multi-resolution methods like the adaptive wavelet transform and its dual, the cosine packet transform, provide a promising alternative. This paper treats the received signal as being made up of four components – tonals, transients, time/frequency transients, and spectrally smooth noise. We introduce an algorithm (ASC) that performs the automated detection and extraction of these four different types of signals. The ASC algorithm has already found applications in the processing of towed array data, humpback whale song and autonomous recorded acoustic datasets collected in Singapore waters. Introduction. The main objective of this paper is to show that standard Wavelet packet and Cosine |
|---|