Acquisition, Tracking and Pointing Technologies

Current acquisition, tracking and pointing (ATP) technology typically incorporates algorithms, such as the Kalman filter, known to be poorly suited to the tracking of nonlinearly moving targets. Nonlinearities may be due to target motion (e.g., agile target) but targeting platform induced jitter also manifests as target nonlinearity. Current and future mobile MDA assets such as ALTB and ABIR will assuredly impart jitter to an image stream that must be compensated to ensure accurate ATP. Algorithms well suited to nonlinear targeting strategies, such as the particle filter algorithm (PFA), can be used to broaden the ATP application space but are computationally expensive (i.e., slow). Polaris Sensor Technologies proposes to advance the PFA by incorporating a filtering component based on the Scale Invariant Feature Transform (SIFT) coupled to a Random Sample Consensus (RANSAC) algorithm. These algorithms are easily incorporated into the PFA framework and promise to improve both speed and accuracy of ATP tasks. The SIFT-RANSAC will remove jitter induced nonlinearities while the PFA will account for target nonlinearities. The combination has direct relevance to a number of MDA programs including legacy and future interceptors such as THAAD, SM3-IIB, and airborne surveillance assets such as ALTB and ABIR.