Output details
11 - Computer Science and Informatics
Imperial College London
Probabilistic Tracking of Affine-Invariant Anisotropic Regions
<23>Effective feature tracking is a fundamental problem in computer vision. Its use for surgical guidance is challenging and affected by the paucity of reliable anatomical features, tissue deformation and inter-reflecting lighting conditions. This paper proposes a new probabilistic framework to track affine-invariant anisotropic regions with comprehensive comparison against the current state-of-the-art. IEEE TPAMI is the top journal in computer vision and pattern recognition and the paper received Best Presentation Award at the 2013 Rank Prize Symposium. It also contributed to the Micro-IGES grant (£2.4M) funded by Wellcome/DoH Health Challenge Fund for developing a novel vision guided microsurgery robot.