The image registration inevitably involves interpolation problems to estimate gray values of the image at positions other than the grid points. In this paper interpolation induced effects on the accuracy of registered results are investigated by using a mutual information measure. Three interpolation techniques, namely nearest neighbor interpolation, bilinear interpolation and partial volume interpolation, are investigated under some elaborately designed experiments. The rigid transformation is limited and the simplex search strategy is adopted. The effects of histogram bin numbers and the subsampling rates are also investigated. Experimental results demonstrate that the partial volume interpolation shows much more smoothness of the mutual information function than nearest neighbor and bilinear interpolations. It is much more robust to improve the accuracy than the other two interpolations. The suitable decrease of histogram bin numbers can further smooth the mutual information function and increase the accuracy of the registration results. The decrease of the subsampling rates greatly saves the computation time, and it results in little losses of the accuracy of the final registered results, when using the partial volume interpolation.
[1]
Jan Modersitzki,et al.
Numerical Methods for Image Registration
,
2004
.
[2]
A. Ardeshir Goshtasby,et al.
2-D and 3-D Image Registration: for Medical, Remote Sensing, and Industrial Applications
,
2005
.
[3]
William H. Press,et al.
Numerical Recipes in C, 2nd Edition
,
1992
.
[4]
Guy Marchal,et al.
Multimodality image registration by maximization of mutual information
,
1997,
IEEE Transactions on Medical Imaging.
[5]
Max A. Viergever,et al.
Interpolation Artefacts in Mutual Information-Based Image Registration
,
2000,
Comput. Vis. Image Underst..
[6]
Guy Marchal,et al.
Multi-modality image registration by maximization of mutual information
,
1996,
Proceedings of the Workshop on Mathematical Methods in Biomedical Image Analysis.
[7]
Paul A. Viola,et al.
Alignment by Maximization of Mutual Information
,
1997,
International Journal of Computer Vision.