S.M. Viitanen (Finland)
Fractal Approaches, Ship Recognition, Passive Recogni tion, Underwater Target.
Due to the complexity of the radiated sound of marine ves sels, both surfaced and submerged, the regularities essen tial to recognition of the vessels are very difficult to obtain. This paper studies the effectiveness of fractal based approaches to the passive recognition of surface and underwater targets. Fractal approaches proposed include wavelet analysis, fractal dimension analysis, and fractional Brownian motion based analysis. For a comparison, a tra ditional feature extraction method based on average spec trum is also included. The methods are tested with sound data from two surface ships and a submarine, along with data from a towed sound source radiating submarine-like noise. Experimental results show that fractal approaches can be very effective in marine target recognition, and may be used to augment the existing recognition methods.
Important Links:
Go Back
