Based on B-splines non-rigid registration method for atmospheric turbulence degraded image

This paper proposes a B-splines non-rigid registration method which applies to atmospheric turbulence image, and then the method applies B-splines registration to multi-frames images and puts forward a two-step control point adjustment method in the process of establishing. B-splines controls point grid that makes image registration from “coarse” to “accuracy” in two steps. As a consequence, experiments show that this method can simulate the details of turbulence degraded image. Meanwhile, it achieves better registration results.

[1]  Xiang Zhu,et al.  Image reconstruction from videos distorted by atmospheric turbulence , 2010, Electronic Imaging.

[2]  Masatoshi Okutomi,et al.  Super-resolution from image sequence under influence of hot-air optical turbulence , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Wei Su,et al.  Research on Cross-Correlative Blur Length Estimation Algorithm in Motion Blur Image , 2016, J. Adv. Comput. Intell. Intell. Informatics.

[4]  Daniel Rueckert,et al.  Nonrigid registration using free-form deformations: application to breast MR images , 1999, IEEE Transactions on Medical Imaging.

[5]  Fred L. Bookstein,et al.  Principal Warps: Thin-Plate Splines and the Decomposition of Deformations , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Andrew Lambert,et al.  Image Restoration of Images Obtained by Near-Horizontal Imaging through the Atmosphere , 2002 .

[7]  Dah-Jye Lee,et al.  Digital Registration Techniques For Sequential Fundus Images , 1988, Optics & Photonics.

[8]  C. D. Kuglin,et al.  The phase correlation image alignment method , 1975 .

[9]  Michael Unser,et al.  Fast parametric elastic image registration , 2003, IEEE Trans. Image Process..

[10]  S. Farsiu,et al.  Constrained, globally optimal, multi-frame motion estimation , 2005, IEEE/SP 13th Workshop on Statistical Signal Processing, 2005.