Target detection in LADAR range images using shadow analysis

In this paper a knowledge base driven technique is presented to detect targets based on the geometric properties of the "shadows" cast by the targets in laser radar (LADAR) range images. The geometric properties of the shadows are relatively easy to compute and are less susceptible to target obscuration. In this technique synthetic images of targets are utilized to populate the knowledge base. The validity of the approach is demonstrated by detecting an armed personnel carrier and a tank utilizing data from captive flight tests.

[1]  Heggere S. Ranganath,et al.  Object detection using pulse coupled neural networks , 1999, IEEE Trans. Neural Networks.

[2]  Mark R. Stevens,et al.  Multisensor occlusion reasoning , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[3]  Jens-Rainer Ohm,et al.  Shape retrieval with robustness against partial occlusion , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[4]  Robert B. Fisher,et al.  Reconstruction of surfaces behind occlusions in range images , 2001, Proceedings Third International Conference on 3-D Digital Imaging and Modeling.

[5]  Peng-Seng Toh,et al.  Recognition of partially occluded object , 1993, Proceedings of TENCON '93. IEEE Region 10 International Conference on Computers, Communications and Automation.