Tactical multisensor fusion (TMSF): an SAIC white paper

Abstract

A software application for correlating and fusing information products from multiple dissimilar sensors is presented. The Tactical Multi Sensor Fusion (TMSF) system is a C ++ object oriented application implementing data correlation and fusion algorithms which provide decision aids for identifying, locating, and determining the status of processing equipment within suspected weapons of mass destruction (WMD) sites. The TMSF system also provides valuable information for assessing weapon delivery accuracy and effectiveness. The TMSF architecture is designed to handle an arbitrary number of sensor and information types, by creating new C ++ class objects. Three specific sensor data types are currently used in the correlation and fusion process. When deployed against suspected WMD sites in a real world scenario, the selected sensor suite provides site characterization information during preattack covert surveillance through post-attack battle damage assessment (BDA). The three sensors are: 1) Tactical Unattended Ground Sensor (TUGS): Seismic/Acoustic in-ground sensors which, when used in sufficient numbers, provides equipment identification, location, and activity with the appropriate error estimates. 2) Forward Looking Infra Red (FLIR): The recorded images from an attack aircraft FLIR pod are used to determine weapon location, time, criticality of detonation, and post detonation venting. 3) Weapon Borne Sensors (WBS): A weapon borne sensor is used to relay information with respect to the construction of the WMD site under attack. This primarily consists of the number and types of layers of resistance met by the weapon as it penetrates the buried facility prior to detonation. All sensor data are interpreted from flat ASCII files as a list of individual event reports, and are correlated with respect to location, ID, and time. The correlation process groups event reports for like-events as reported by different sensors. The correlated groups are then fused into composite site characterization objects (SCO) using Bayes maximum likelihood based algorithms to obtain a fused location, ID, and time estimate. An image overlay is used to locate each identified processing equipment on an overhead site image. This composite view of the fusion results gives the Measurement and Signals Intelligence (MASINT) specialist, intelligence analyst, or military targeteer the capability to interpret the fusion results without the need for tedious, accident prone, numerical data analysis. TMSF reads and displays the National Imagery Transmission Format (NITF 2.0) image files as well as other commercially available formats.