Share Email Print

Proceedings Paper

Anomaly detection using range profile and intensity signatures
Author(s): Stephen Cain; Brian Deas
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Anomaly detection methods applied to LAser Detection And Ranging (LADAR) data have traditionally utilized intensity and range as indicators. The proposed method for anomaly detection presented in this paper uses information about the range profile of the target to determine if the target is anomalous. The range profile is related to the range depth of the target surface relative to the illuminating wave of the LADAR system. If the target surface is slanted for instance, the range to the target will vary as a function of position within the instantaneous field of view of the LADAR receiver optics. This variation of range can be indicative of the presence of an anomaly within the scene viewed by the system. The ability to detect range variation from a returned LADAR pulse is a feature of the Normalized VAriable Shape (NOVAS) correlator. The algorithm computes both the range to the target and the shape of the pulse adaptively using an assumed mathematical form for the shape of the pulse. In the cases examined in this paper a Gaussian pulse shape is assumed for the pulse returning from the target. The NOVAS algorithm estimates the standard deviation of the Gaussian as well as the range associated with the shift in its position in time. The NOVAS algorithm achieves its ability to discern both the range to the target and the shape of the returning pulse through the use of the Pearson's product coefficient. This normalized correlation operation searches over both range and pulse shape parameter to achieve the highest correlation value between the pulse model and the reflected pulse measurement. The range depth of the target can be inferred from the returning pulse width as the pulse shape reflected from the target is the result of a convolution between the pulse fired at the target and the range profile.

Paper Details

Date Published: 22 April 2010
PDF: 9 pages
Proc. SPIE 7687, Active and Passive Signatures, 768703 (22 April 2010); doi: 10.1117/12.850351
Show Author Affiliations
Stephen Cain, Air Force Institute of Technology (United States)
Brian Deas, Air Force Institute of Technology (United States)

Published in SPIE Proceedings Vol. 7687:
Active and Passive Signatures
G. Charmaine Gilbreath; Chadwick T. Hawley, Editor(s)

© SPIE. Terms of Use
Back to Top