Diver breathing sounds can be used as a characteristic for the passive detection of divers. This work introduces an approach for detecting the presence of a diver based on diver breathing sounds signals. An underwater channel model for passive diver detection is built to evaluate the impacts of acoustic energy transmission loss and ambient noise interference. The noise components of the observed signals are suppressed by spectral subtraction based on block-based threshold theory and smooth minimal statistic noise tracking theory. Then the envelope spectrum features of the denoised signal are extracted for diver detection. The performance of the proposed detection method is demonstrated through experimental analysis and numerical modeling.

No abstract

The acoustic technology has a great potential to measure the concentration of particles in water which is one of the parameters for understanding ocean ecology. To achieve this purpose, the forward acoustic propagation model has been considered as a method to obtain concentration and particle size. The Gaussian lognormal distribution was used to describe the particle matters condition in suspended water. Base on the model, a nonlinear inversion process took values of attenuation into calculation and get parameters of concentration, average radius and variance.

No abstract

No abstract