Threat Assessment Using Context-Based Tracking in a Maritime Environment Jemin George John L. Crassidis Tarunraj Singh Graduate Student Professor Professor jgeorge3@buffalo.edu johnc@buffalo.edu tsingh@buffalo.edu Department of Mechanical & Aerospace Engineering University at Buffalo, State University of New York, Amherst, NY 14260–4400 Abstract – The main objective of this work is to model and exploit available maritime contextual information to provide a hypothesis on suspicious vessel maneuvers. This concept involves utilizing the L1 tracking to perform L2/L3 data fusion, i.e. refinement and assessment for situations and threats. A new context-based tracker known as the Con- Tracker is developed. The purpose of the ConTracker is to incorporate the contextual information into a traditional α − β tracker in such a way so that it provides a repeller or an attractor characteristics to a specific region of inter- est. Any behavior of the vessel that is inconsistent with the repeller or the attractor characteristics of the current vessel location would be classified as suspicious. Such an incon- sistent vessel behavior would be directly indicated by a high measurement residual which may be used to estimate an accurate process noise covariance using a multiple-model adaptive estimator. Based on the rate of change of the esti- mated process noise covariance values, an L2/L3 hypothe- sis generator red-flags the target vessel. Simulation results indicate that the context based tracking enhances the relia- bility of erratic maneuver detection. Keywords: Trafficability, ConTracker, MMAE, L2/L3 fu- sion, α − β tracker. 1 Introduction Traditional tracking algorithms heavily rely on target model and observations but do not exploit local informa- tion. Though these approaches work well for some targets, they often fail to account for the movements of intelligent objects. Advancement of complex tracking schemes suggest that increasing the amount of information included in the al- gorithm can improve the quality of the tracking process. A terrain-based tracking approach which accounts for the ef- fects of terrain on target speed and direction of movement is presented in Ref. [1]. It has been shown that the incorpora- tion of local contextual information such as the terrain data can significantly improve the tracker performance [2]. In recent years, researchers have explored the overt use of con- textual information for improving state estimation in ground target tracking by incorporating this information into the tracking algorithm as a potential field to provide a repeller or an attractor characteristic to a specific region of interest [3]. In Ref. [4], the local contextual information, termed traffi- cability, incorporates local terrain slope, ground vegetation and other factors to put constraints on the vehicle’s max- imum velocity. Simulation results given in Ref. [4] show that the use of trafficability can improve estimate accuracy in locations where the vehicle path is influenced by terrain features. The main goal of this work is to exploit available mar- itime information to provide a hypothesis on suspicious boat movements. For example, it is desired to “red-flag” a boat that approaches a restricted high value unit area. Also, a vessel that is erratically zigzagging across a marked ship- ping channel may also be red-flagged for suspicious activ- ity. The process to provide a hypothesis of this notion is de- picted in Fig. 1. This concept involves exploiting the math- ematical rigorous approaches of L1 tracking in an L2/L3 situation and threat refinement and assessment scheme (see Ref. [5] for Joint Directors of Laboratories’ description of the various data fusion levels). The proposed methodology consists of three main components; a context-based tracker called ConTracker, a Multiple Model Adaptive Estimator (MMAE), and a hypothesis generator. Measurements Residual Motion Estimation ConTracker Adaptive ID Process MMAE L2/L3 Hypothesis Generator State Estimates Contextual Information A-priori Hypothesis Figure 1: System Flowchart The ConTracker (for Context-based Tracker) combines 12th International Conference on Information Fusion Seattle, WA, USA, July 6-9, 2009 978-0-9824438-0-4 ©2009 ISIF 187