This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Online Learning of Correspondences between Images
Jan. 2013 (vol. 35 no. 1)
pp. 118-129
M. Felsberg, Dept. of Electr. Eng., Linkoping Univ., Linkoping, Sweden
F. Larsson, Dept. of Electr. Eng., Linkoping Univ., Linkoping, Sweden
J. Wiklund, Dept. of Electr. Eng., Linkoping Univ., Linkoping, Sweden
N. Wadstromer, Div. of Inf. Syst., FOI Swedish Defence Res. Agency, Linkoping, Sweden
J. Ahlberg, Termisk Systemteknik AB, Linkoping, Sweden
We propose a novel method for iterative learning of point correspondences between image sequences. Points moving on surfaces in 3D space are projected into two images. Given a point in either view, the considered problem is to determine the corresponding location in the other view. The geometry and distortions of the projections are unknown, as is the shape of the surface. Given several pairs of point sets but no access to the 3D scene, correspondence mappings can be found by excessive global optimization or by the fundamental matrix if a perspective projective model is assumed. However, an iterative solution on sequences of point-set pairs with general imaging geometry is preferable. We derive such a method that optimizes the mapping based on Neyman's chi-square divergence between the densities representing the uncertainties of the estimated and the actual locations. The densities are represented as channel vectors computed with a basis function approach. The mapping between these vectors is updated with each new pair of images such that fast convergence and high accuracy are achieved. The resulting algorithm runs in real time and is superior to state-of-the-art methods in terms of convergence and accuracy in a number of experiments.
Index Terms:
vectors,image sequences,learning (artificial intelligence),matrix algebra,optimisation,basis function approach,correspondence online learning,point correspondence iterative learning,image sequences,3D space,correspondence mappings,global optimization,fundamental matrix,perspective projective model,point-set pair sequence,general imaging geometry,Neyman chi-square divergence,channel vectors,Vectors,Cameras,Estimation,Geometry,Channel estimation,Three dimensional displays,Accuracy,surveillance,Online learning,correspondence problem,channel representation,computer vision
Citation:
M. Felsberg, F. Larsson, J. Wiklund, N. Wadstromer, J. Ahlberg, "Online Learning of Correspondences between Images," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 35, no. 1, pp. 118-129, Jan. 2013, doi:10.1109/TPAMI.2012.65
Usage of this product signifies your acceptance of the Terms of Use.