SBIR-STTR Award

Detection, tracking, and identification of ballistic missiles can benefit significantly from improved correlation of the tracks and detections. Track and detections are coming from the optical sensors and the radar sensors for further fusion of data into the single picture of the situation. However, available association methods are not nearly robust enough. Several radar, EO, and IR sensor characteristics reduce the ability to match tracks correctly. Range and azimuth resolution, scale distortion, relief displacement (foreshortening and layover), and sensor shadow are among them. The objective of this proposal is to develop advanced, innovative, robust, real-time algorithms and software for the integration of passive or active EO sensor tracks or detections with radar generated tracks or detections. The solution will use metric data and features, provide a measure of confidence, identify clusters of tracks when objects are indistinguishable, and can be implemented in a centralized or distributed architecture. The main novelty of the proposed EAD algorithm is in the use of robust and affine invariant structural relations built on the features for accurate correlation.

Keywords:
Track Correlation, Sensor Fusion, Data Fusion, Multi-Sensor

Detection, tracking, and identification of ballistic missiles can benefit significantly from improved correlation of the tracks and detections. Tracks and detections are coming from the optical sensors and the radar sensors for further fusion of data into the single picture of the situation. However, available association/correlation methods are not nearly robust enough. Several radar, EO, and IR sensor characteristics reduce the ability to match tracks correctly. Range and azimuth resolution, scale distortion, relief displacement (foreshortening and layover), and sensor shadow are among them. The objective of this proposal is to develop advanced, innovative, robust, real-time algorithms and software for the integration of passive or active EO sensor tracks or detections with radar generated tracks or detections. The solution uses metric data and features, provides a measure of confidence, identifies clusters of tracks when objects are indistinguishable, and can be implemented in a centralized or distributed architecture. The main novelty of the proposed EAD algorithm is in the use of robust and affine invariant structural relations built on the features for accurate correlation.

Keywords:
Track Correlation, Track Association, Fusion, Eo/Ir, Radar, Sensor Netting, Invariance, Robustness