In conjunction with the Bremerton Detachment of the Carderock Division of the Naval Surface Warfare Center, we will provide algorithms for tests on signals from underwater transient signals. These signals have already been found to be well-represented by positive time-frequency distributions. Our goal is to help this Navy Lab (and other labs as directed by the ONR) with extending these results to the use of the various auditory representations described in Thrust area I, and design of automated classification and detection systems based on the algorithms described in Thrust area II. The NSWC has extensive data bases which can be used to train and test proposed algorithms. As can be seen from the attached letter, NSWC is very interested in taking this role to help this proposed center's researchers direct their algorithms to these Navy problems.
It should also be pointed out that there is a strong underlying similarity between the work in the above sections A and C, and this underwater detection application. For example, it has been previously found that vibration transients in machining carry the richest information about tool-edge breakdown [.Atlas Bernard 1995.]. Thus, many of the algorithm developments for the above section C could fit, with slight modification, the problems inherent in classification of underwater transients.