Cad-based Target Identiication in Range, Ir and Color Imagery Using On-line Rendering and Feature Prediction Aaliation of Authors Contact Information

Results for a mutlisensor CAD-based object recognition system are presented in the context of Automatic Target Recognition using nearly boresight-aligned range, IR and color sensors. The system is shown to identify targets in test suite of 35 image triples. This suite includes targets at low resolution, unusual aspect angles, and partially obscured by terrain. The key concept presented in this work is that of using on-line rendering of CAD-models to support an iterative predict, match and re ne cycle. This cycle optimizes the match subject to variability both in object pose and sensor registration. An occlusion reasoning component further illustrates the power of this approach by customizing the predicted features to t speci c scene geometry. Occlusion reasoning detects occlusion in the range data and adjusts the features predicted to be visible accordingly.