Image registration based on Genetic Algorithm and weighted feature correspondences

Super resolution is a technique of enhancing image resolution by combining information from multiple images. It is widely applied in fields like camera surveillance, satellite imaging, pattern recognition, etc. One challenging problem of super resolution is its high demand on image registration accuracy. This paper introduces a high accuracy registration approach for the purpose of super resolution. It is invariant to translation, scaling, rotation, and noise, and can be used to automatically obtain the Maximize a Likelihood Estimation (MLE) of image homography (registration result) using information only contained within the images themselves. An effective Genetic Algorithm based approach is used to filter out all the mismatches. Comparison with RANSAC and Keren's method will be given to prove the effectiveness of the proposed method.

[1]  Bernhard P. Wrobel,et al.  Multiple View Geometry in Computer Vision , 2001 .

[2]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[3]  P. Anandan.  Model based techniques for image registration and three-dimensional scene analysis from image sequences , 1993, Proceedings of 27th Asilomar Conference on Signals, Systems and Computers.

[4]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[5]  M. Okutomi,et al.  Robust and Accurate Image Registration with Pixel Selection , 2006, 2006 IEEE International Symposium on Signal Processing and Information Technology.

[6]  Michael Unser,et al.  A pyramid approach to subpixel registration based on intensity , 1998, IEEE Trans. Image Process..

[7]  Matthew A. Brown,et al.  Invariant Features from Interest Point Groups , 2002, BMVC.

[8]  B. S. Manjunath,et al.  A contour-based approach to multisensor image registration , 1995, IEEE Trans. Image Process..

[9]  B. N. Chatterji,et al.  An FFT-based technique for translation, rotation, and scale-invariant image registration , 1996, IEEE Trans. Image Process..

[10]  Stephen J. Roberts,et al.  Bayesian Image Super-resolution, Continued , 2006, NIPS.

[11]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[12]  Shmuel Peleg,et al.  Image sequence enhancement using sub-pixel displacements , 1988, Proceedings CVPR '88: The Computer Society Conference on Computer Vision and Pattern Recognition.

[13]  Gabriel Taubin,et al.  Estimation of Planar Curves, Surfaces, and Nonplanar Space Curves Defined by Implicit Equations with Applications to Edge and Range Image Segmentation , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[14]  Amir Averbuch,et al.  FFT based image registration , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[15]  Michael T. Orchard,et al.  A fast direct Fourier-based algorithm for subpixel registration of images , 2001, IEEE Trans. Geosci. Remote. Sens..

[16]  Stephen J. Roberts,et al.  Overcoming Registration Uncertainty in Image Super-Resolution: Maximize or Marginalize? , 2007, EURASIP J. Adv. Signal Process..

[17]  Jianya Gong,et al.  POCS Super-Resolution Sequence Image Reconstruction Based on Improvement Approach of Keren Registration Method , 2006, Sixth International Conference on Intelligent Systems Design and Applications.

[18]  Michael Unser,et al.  A pyramid approach to sub-pixel image fusion based on mutual information , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[19]  Jianya Gong,et al.  An improvement approach based on keren sub-pixel registration method , 2006, 2006 8th international Conference on Signal Processing.