Pose estimation from airborne video sequences using a structural approach for the construction of homographies and fundamental matrice

A structural knowledge-based search method is utilized for the estimationof geometric transforms from airborne video sequences. Examples areprojective planar homographies and constraints such as the fundamentalmatrix. These estimations are calculated from correspondences ofinterest points between two images. Different approaches are discussedto cope with the problem of outlier-correspondences. To ensure any-timeperformance the search process is implemented in a data-driven productionsystem. The pose estimation from planar homographies is comparedto estimations from fundamental matrices. A fusion of both approachesis proposed. The image processing is performed by bottom-up structuralanalysis using an assessment-driven control. Examples are from thethermal spectral domain.     «

A structural knowledge-based search method is utilized for the estimationof geometric transforms from airborne video sequences. Examples areprojective planar homographies and constraints such as the fundamentalmatrix. These estimations are calculated from correspondences ofinterest points between two images. Different approaches are discussedto cope with the problem of outlier-correspondences. To ensure any-timeperformance the search process is implemented in a data-driven productionsystem. The...     »