摘要: The basic limitations of the standard appearance-based matching methods using eigenimages are nonrobust estimation coefficients and inability to cope with problems related outliers, occlusions, varying background. In this paper we present a new approach which successfully solves these problems. major novelty our lies in way determined. Instead computing by projection data onto eigenimages, extract them robust hypothesize-and-test paradigm subsets image points. Competing hypotheses then subject selection procedure based on Minimum Description Length principle. enables us not only reject outliers deal occlusions but also simultaneously use multiple classes eigenimages.
elsevier.com
sci-hub.se 参考文章(34)
A. Bab-Hadiashar, D. Suter, Optic flow calculation using robust statistics computer vision and pattern recognition. pp. 988- 993 ,(1997) , 10.1109/CVPR.1997.609448
Markus Stricker, Aleš Leonardis, ExSel++: A General Framework to Extract Parametric Models computer analysis of images and patterns. pp. 90- 97 ,(1995) , 10.1007/3-540-60268-2_284
Colin R. Reeves, Modern heuristic techniques for combinatorial problems ,(1993)
R.P.N. Rao, Dynamic appearance-based recognition computer vision and pattern recognition. pp. 540- 546 ,(1997) , 10.1109/CVPR.1997.609378
J. Edwards, H. Murase, Appearance matching of occluded objects using coarse-to-fine adaptive masks computer vision and pattern recognition. pp. 533- 539 ,(1997) , 10.1109/CVPR.1997.609377
S. Yoshimura, T. Kanade, Fast template matching based on the normalized correlation by using multiresolution eigenimages intelligent robots and systems. ,vol. 3, pp. 2086- 2093 ,(1994) , 10.1109/IROS.1994.407577
A. Leonardis, H. Bischof, Dealing with occlusions in the eigenspace approach computer vision and pattern recognition. pp. 453- 458 ,(1996) , 10.1109/CVPR.1996.517111
CE Shennon, Warren Weaver, A mathematical theory of communication Bell System Technical Journal. ,vol. 27, pp. 379- 423 ,(1948) , 10.1002/J.1538-7305.1948.TB01338.X
Aleš Leonardis, Horst Bischof, An efficient MDL-based construction of RBF networks Neural Networks. ,vol. 11, pp. 963- 973 ,(1998) , 10.1016/S0893-6080(98)00051-3
Elvezio M. Ronchetti, Peter J. Rousseeuw, Werner A. Stahel, Frank R. Hampel, Robust statistics: the approach based on influence functions ,(1986)