Signal Processing

Sensor Data Fusion

Information is King. Increased sensor diversity can provide dramatic performance gains in scene characterization, event detection, and target recognition. Such diversity can be realized with multiple look-geometries and/or multiple sensing modalities (e.g. hyperspectral + radar). As sensor diversity grows, the challenge is to integrate information critical to the mission application while simultaneously filtering out irrelevant information. SIG has developed new approaches for intelligent multi-sensor data integration and decision inference. SIG uses mathematical techniques such as variational methods and Bayesian graph-theory that optimize decisions based on impartial sensor data. SIG’s Active Learning technology can guide sensor management and deployment to make better decisions while minimizing cost/risk.

Radar Signal Processing

Radar is a key sensor technology due to its ability to operate under weather and other environmental conditions inhospitable to other sensors. We at SIG have the ability to process and analyze raw radar data from various radar modalities. Our experience includes High Range Resolution (HRR) signature processing, Synthetic Aperture Radar (SAR) image formation, motion compensation techniques (MoComp), advanced Radio Frequency Interference (RFI) mitigation, adaptive beamforming, and Space Time Adaptive Processing (STAP) for Ground Moving Target Indication (GMTI) modes. Additionally, our capabilities include radar system simulation and electromagnetic modeling, such as this SAR image of a set of modeled capped-cylinder fiducial structures. Our radar expertise allows us to exploit the unique features apparent in the radar spectrum for classification of targets of interest.