Accession Number : ADA309319
Title : Nonparametric Multisensor Track Initiation Method for Nonuniform Nonstationary Environments Part II: A Comparison of M/N and ANMSTI Track Initiation Techniques Applied to the Wallops Island Multisensor Integration Experiment.
Descriptive Note : Final rept.,
Corporate Author : NAVAL SURFACE WARFARE CENTER DAHLGREN DIV VA
Personal Author(s) : Pawlak, Robert J. ; Stapleton, Ron A.
PDF Url : ADA309319
Report Date : JUL 1995
Pagination or Media Count : 21
Abstract : This follow-on report compares the performance of the adaptive nonparametric multisensor track initiation (ANMSTI) algorithm to that of conventional M out of N (MIN) track initiation algorithms, for a test set composed of concurrent radar and infrared search and track (IRST) data. To present a fair comparison of the various techniques, both false track initiation rates and track initiation times are analyzed in this report. While the analysis shows that some of the M/N techniques provide better track initiation times than the ANMSTI approach, some of these same techniques have false track initiation rates that in cluttered environments render them almost useless. Furthermore, it is shown that the ANMSTI algorithm is capable of regulating its false track initiation rate, whereas the M/N algorithms cannot. When comparing the time to first track initiation of single sensor and dual sensor M/N techniques, most of the benefits of using a dual sensor M/N track initiation criteria are shown to accrue when the track initiation criterion is relatively strict (i.e., usually when false alarm rates from the sensors are high).
Descriptors : *TARGET ACQUISITION, *INFRARED TRACKING, ALGORITHMS, COMPARISON, CLUTTER, FALSE ALARMS, INTEGRATION, SEARCHING, MULTISENSORS, TEST SETS.
Subject Categories : Infrared Detection and Detectors
Distribution Statement : APPROVED FOR PUBLIC RELEASE