The increasing needs of registered images, usually picked up from diverse moving sources of deformable nature, have arisen in diverse fields and have focused the attention of researchers in the last years. The critical step involved within image registration use to be image matching. This paper proposes a robust method that faces the ill-posed nature of this image matching process, taking into account the image rotation effect. The matching ambiguity is described by a fuzzy parametric model that includes estimated relative rotation, and, finally, a spatially non-uniform fuzzy interpolation is used to translate the parametric info into a set of matching field vectors. The method obtains the spatial matching between the two images in a global spatial extent and with sub-pixel accuracy. Results of the method on real images and high non-rigid artificial deformation and rotation prove the validity of the approach.
[1]
Luis Weruaga,et al.
Frequency domain formulation of active parametric deformable models
,
2004,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[2]
R. Verdu,et al.
Non-rigid motion estimation based on fuzzy models
,
2002,
2002 14th International Conference on Digital Signal Processing Proceedings. DSP 2002 (Cat. No.02TH8628).
[3]
C. D. Kuglin,et al.
The phase correlation image alignment method
,
1975
.
[4]
L. Weruaga,et al.
Estimating volumetric motion in human thorax with parametric matching constraints
,
2003,
IEEE Transactions on Medical Imaging.
[5]
Thierry Pun,et al.
Rotation, scale and translation invariant digital image watermarking
,
1997,
Proceedings of International Conference on Image Processing.