Ship sensor Fusion for Target Identification and Classification (SIFTIC)

Stottler Henke proposes the Ship sensor Fusion for Target Identification and Classification (SIFTIC), which will combine three existing projects: ExPATSS, IR/Radar Fusion, and FusionCommand to fuse data from multiple sensors of different phenomenologies to classify ships. SIFTIC will collect type agnostic sensor data, including Imaging Infrared (IIR), Inverse Synthetic Aperture Radar (ISAR), electro-optical (EO) images, electronic support measures (ESM), light detection and ranging (LIDAR), and automatic identification system (AIS) data. The interface with many of these types of sensors has already been developed by our subcontractor Global Technical Systems (GTS) via their FusionCommand. The collected data will be fused using feature extraction, Bayesian networks, case-based reasoning, and multiple hypothesis reasoning. The fusion of IIR and ISAR data has already been developed in Stottler Henkes IR/Radar Fusion project, and will be expanded to other possible sensor types as well. Finally, the fused data sets will be classified using computer vision techniques and deep learning. These techniques have all been implemented and applied to IIR and ISAR images in Stottler Henkes ExPATSS project. We also have the opportunity to test on real data collected by GTS and Raytheon as well as lab data of accurate ship models from UDRIs Mumma Radar Lab.