Title :
The neurally-inspired contact estimator (NICE)
Author :
DeAngelis, Christopher M. ; Whitney, James E.
Author_Institution :
Naval Undersea Warfare Center, Newport, RI, USA
fDate :
28 Sep-1 Oct 1998
Abstract :
The neurally-inspired contact estimator (NICE) is a signal processing system designed to perform target motion analysis, that is, source localization and tracking given a series of bearing observations. NICE is implemented using a custom neural network, which incorporates the characteristics of an object´s motion parameters into the network weights. In this paper the relative fit between NICE and conventional probability theory is analyzed
Keywords :
correlation methods; direction-of-arrival estimation; motion estimation; neural nets; probability; sensor fusion; sonar signal processing; sonar tracking; target tracking; time series; assimilation subsystem; bearing observations; conventional probability theory; correlation; custom neural network; data fusion; hydrophone array; network weights; neurally-inspired contact estimator; relative fit; signal processing system; source localization; spatial simulation; target motion analysis; time series; tracking; virtual retina; Machine vision; Motion analysis; Motion estimation; Neural networks; Phased arrays; Radar tracking; Retina; Signal processing; Sonar equipment; Target tracking;
Conference_Titel :
OCEANS '98 Conference Proceedings
Conference_Location :
Nice
Print_ISBN :
0-7803-5045-6
DOI :
10.1109/OCEANS.1998.726362
