In this paper, an outlier elimination algorithm for ellipse/ellipsoid fittingis proposed. This two-stage algorithm employs a proximity-based outlierdetection algorithm (using the graph Laplacian), followed by a model-basedoutlier detection algorithm similar to random sample consensus (RANSAC). Thesetwo stages compensate for each other so that outliers of various types can beeliminated with reasonable computation. The outlier elimination algorithmconsiderably improves the robustness of ellipse/ellipsoid fitting asdemonstrated by simulations.
展开▼