Automated visual inspection systems have the potential to improve space situational awareness capabilities. However, satellite inspection systems are subject to harsh lighting conditions and bandwidth constraints which necessitate a unique approach. This thesis presents two approaches to automated change detection.;The first method calculates a planar homography which relates the two images, so that the viewpoint of one image may be warped to the viewpoint of the other. The second method estimates the pose of the object in the image, so that a synthetic comparison image can be generated from a pre-existing model of the object. Each algorithm was simulated under varying lighting and pose conditions.
展开▼
