The signal-to-noise ratio of human divers

A challenging problem in underwater acoustics is the detection and tracking of human divers. In order to assess the suitability of different tracking methods on this problem, an evaluation of the signal-to-noise ratio (SNR) is required. This paper investigates the SNR of a human diver using two approaches. The first approach is a moment-based target power estimator which uses the amplitudes of measurements validated by the tracking method. The second approach is by means of the receiving operating characteristic (ROC). The integrated Probabilistic Data Association (IPDA) is used to track the diver. To the best of our knowledge, this is the first well-documented implementation of the IPDA on real data.