Measuring the quality of hypotheses in model-based recognition
Model-based recognition methods generally search for geometrically consistent pairs of model and image features. The quality of an hypothesis is then measured using some function of the number of model features that are paired with image features. The most common approach is to simply count the number of pairs of consistent model and image features. However, this may yield a large number of feature pairs, due to a single model feature being consistent with several image features and vice versa. A better quality measure is provided by the size of a maximal bipartite matching, which eliminates the multiple counting of a given feature. Computing such a matching is computationally expensive, but under certain conditions it is well approximated by the number of distinct features consistent with a given hypothesis.
- 2.Cass, T.A., 1990, “Feature Matching for Object Localization in the Presence of Uncertainty”, MIT Artificial Intelligence Laboratory Memo no. 1133.Google Scholar
- 3.Grimson, W.E.L. and D.P. Huttenlocher, 1990, “On the Verification of Hypothesized Matches in Model-Based Recognition”, Proceedings of the First European Conference on Computer Vision, Lecture Notes in Computer Science No. 427, pp. 489–498, Springer-Verlag.Google Scholar
- 4.Grimson, W.E.L. & T. Lozano-Pérez, 1987, “Localizing overlapping parts by searching the interpretation tree,” IEEE Trans. PAMI 9(4), pp. 469–482.Google Scholar
- 5.Huttenlocher, D.P. and S. Ullman, 1990, “Recognizing Solid Objects by Alignment with an Image”, Intl. Journal of Computer Vision, vol. 5, no. 2, pp. 195–212.Google Scholar
- 6.Huttenlocher, D.P. and Kedem, K., 1990, “Efficiently computing the Hausdorff distance for point sets under Translation”, proceedings of Sixth ACM Symposium on Computational Geometry, pp. 340–349.Google Scholar
- 7.Papadimitriou, C.H. and K. Steiglitz, 1982, Combinatorial Optimization: Algorithms and Complexity, Prentice-Hall.Google Scholar