We describe a model-based image analysis system that automatically estimates the 3-D orientationvectors of satellites and their subcomponents by analyzing images obtained from a ground-basedoptical surveillance system. We adopt a two-step approach: pose estimates are derived fromcomparisons with a model database and pose refinements are derived from photogrammetricinformation. The model database is formed by representing each available training image by a set ofderived geometric primitives. To obtain fast access to the model database and to increase theprobability of early successful matching, a novel index-hashing method is introduced. An affinepoint-matching method is also introduced for improving system performance on a wide variety ofsatellite shapes. We present recent results, which include our efforts at isolating and estimatingorientation vectors from degraded imagery on a significant database of satellites.
展开▼
