Situation Assessment: An End-to-End Process for the Detection of Objects of Interest

Semiautomatic approaches are developed for wide area situation assessment in near-real-time. The two-step method consists of two granularity levels. The first entity assessment uses a new multi-target tracking (MTT) algorithm (hybridization of Gaussian mixture-Cardinalized probability hypothesis density (GM-CPHD) filter and multiple hypothesis tracker (MHT) with road constraints) on ground moving target indicator (GMTI) data. The situation is then assessed by detecting objects of interest such as convoys with other data types (synthetic aperture radar (SAR), video). These detections are based on Bayesian networks and their credibilistic counterpart.