Research work · 2007

Estimating and Refining Orientation Vectors for Sub-Components of Space Object Imagery

Xun Du, Stanley C. Ahalt, Bruce E. Stribling

Abstract

We describe the improvements and enhancements that we have made to a model-based image analysis system which automatically estimates the orientation vector of satellites and their sub-components. The system, implemented in Khoros, operates on images obtained from a ground-based optical surveillance system. Pose estimates are first derived from comparisons with a model database, after which pose refinements are derived from photogrammetric or geometric information. In order to efficiently access the potentially large model database, which is formed by representing each available training image by a set of derived geometric primitives, a novel index hashing method has been introduced. To further improve system performance on a wide variety of satellite images, affine point matching and Hough Transform techniques are implemented for pose refinement. Finally, we introduce one method for obtaining pose estimates of unknown satellites. We present recent results which include our efforts at i...