Non-rigid Image Registration Based on Low-rank Nyström Approximation and Spectral Feature
-
摘要: 图像非刚性配准在计算机视觉和医学图像有着重要的作用.然而存在的非刚性配准算法对严重扭曲变形的图像配准精度和效率都比较低.针对该问题,提出基于Nystrm低阶近似和谱特征的图像非刚性配准算法.算法首先提取像素的谱特征,并将谱特征与空间特征、灰度特征融合形成具有扭曲不变性的全局谱特征; 然后在微分同胚配准的框架内使用全局谱匹配,确保算法产生的变形场具有光滑性、可逆性、可微性,以提高配准的精度;其次采用Nystrm抽样方法,随机抽取拉普拉斯矩阵的行与列,低阶逼近该矩阵,降低高维矩阵谱分解的时间,从而提高配准的效率;最后提出基于小波分解的多分辨率图像配准方法,进一步提高配准的精度和效率.理论分析和实验结果均表明,该算法的配准精度和配准效率都有明显的提高.Abstract: Non-rigid image registration plays an important role in computer vision and medical image. However, the typical non-rigid registration algorithms for seriously distorted deformation images lead to a poor registration precision and low efficiency. Therefore, we introduce a new non-rigid image registration algorithm based on low-rank Nystrm approximation method and spectral feature. Firstly, we extract the spectral feature of pixel, and combine spatial feature and gray feature to form global spectral feature which is invariant to distortion. Then the spectral match method is used within the diffeomorphic registration framework so that the deformation field generated by the algorithm is smooth, reversible, and differentiable, with an improved registration precision. Secondly, we use Nystrm sampling method to speed up high dimension matrix spectral decomposition by generating a low-rank approximation matrix via randomly selected rows and columns of the Laplace matrix. Finally, we put forward an image registration method based on wavelet decomposition, to improve the accuracy and efficiency of registration. The theoretic analysis and experimental results show that our algorithm can improve registration precision as well as efficiency.
-
Key words:
- Non-rigid registration /
- low-rank Nystr? /
- m approximation /
- spectral feature /
- wavelet decomposition
-
[1] Brown L G. A survey of image registration techniques. ACM Computing Surveys (CSUR), 1992, 24(4): 325-376 [2] Zhang Gui-Mei, Jiang Shao-Bo, Liu Pi-Yu, Zhang Song. Affine registration based on CCCTI and hierarchical clustering. Journal of Image and Graphics, 2013, 18(9): 1074-1084 (张桂梅, 江少波, 刘丕玉, 张松. 融合CCCTI码和谱系聚类的仿射配准. 中国图象图形学报, 2013, 18(9): 1074-1084) [3] Zhang Gui-Mei, Jiang Shao-Bo, Chu Jun. Affine registration based on chord height point and genetic algorithm. Acta Automatica Sinica, 2013, 39(9): 1447-1457 (张桂梅, 江少波, 储珺. 基于弦高点和遗传算法的仿射配准. 自动化学报, 2013, 39(9): 1447-1457) [4] Peng Xiao-Ming, Chen Wu-Fan, Ma Qian. Elastic point registration method based on B-splines. Journal of Image and Graphics, 2007, 12(6): 1079-1085 (彭晓明, 陈武凡, 马茜. 基于B样条的弹性点配准方法. 中国图象图形学报, 2007, 12(6): 1079-1085) [5] [5] Thirion J P. Image matching as a diffusion process: an analogy with Maxwell's demons. Medical Image Analysis, 1998, 2(3): 243-260 [6] [6] Vercauteren T, Pennec X, Perchant A, Ayache N. Symmetric log-domain diffeomorphic registration: a demons-based approach. Medical Image Computing and Computer-Assisted Intervention --- MICCAI 2008. Berlin Heidelberg: Springer 2008. 754-761 [7] [7] Tang T W H, Chung A C S. Non-rigid image registration using graph-cuts. Medical Image Computing and Computer-Assisted Intervention --- MICCAI 2007. Berlin Heidelberg: Springer 2007. 916-924 [8] Yan Yu-Wu, Liu Jin-Mang. Nonrigid medical image registration approach based on game theory. Chinese Journal of Scientific Instrument, 2010, 31(9): 2049-2055 (鄢余武, 刘进忙. 非刚性医学图像的博弈配准方法. 仪器仪表学报, 2010, 31(9): 2049-2055) [9] Wang Jian, Pan Jing-Wei, Yang Xin. Non-rigid registration for myocardial perfusion MR image. Journal of Image and Graphics, 2013, 18(6): 661-668 (王建, 潘静薇, 杨新. 心肌灌注核磁共振图像的非刚性配准. 中国图象图形学报, 2013, 18(6): 661-668) [10] Cobzas D, Sen A. Random walks for deformable image registration. Medical Image Computing and Computer-Assisted Intervention --- MICCAI 2011. Berlin Heidelberg: Springer, 2011. 557-565 [11] Lombaert H, Grady L, Pennec X, Ayache N, Cheriet F. Spectral demons-image registration via global spectral correspondence. Computer Vision-ECCV 2012. Berlin Heidelberg: Springer 2012. 30-44 [12] Lombaert H, Grady L, Pennec X, Ayache N, Cheriet F. Spectral log-demons: diffeomorphic image registration with very large deformations. International Journal of Computer Vision, 2014, 107(3): 254-271 [13] Fowlkes C, Belongie S, Chung F, Malik J. Spectral grouping using the Nystrm method. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(2): 214-225 [14] Talwalkar A, Rostamizadeh A. Matrix coherence and the Nystrm methods. In: Proceedings of the 26th Conference in Uncertainty in Artificial Intelligence. arXiv preprint arXiv: 1004.2008, 2010. [15] Drineas P, Mahoney M W. On the Nystrm method for approximating a gram matrix for improved kernel-based learning. The Journal of Machine Learning Research, 2005, 6: 2153-2175 [16] Liu Li, Su Min. Medical image registration based on wavelet transformation and mutual information. Journal of Image and Graphics, 2008, 13(6): 1171-1176 (刘丽, 苏敏. 基于小波变换和互信息的医学图像配准. 中国图象图形学报, 2008, 13(6): 1171-1176) [17] Vercauteren T, Pennec X, Perchant A, Ayache N. Non-parametric diffeomorphic image registration with the demons algorithm. Medical Image Computing and Computer-Assisted Intervention --- MICCAI 2007. Berlin Heidelberg: Springer, 2007. 319-326 [18] Klein A, Andersson J, Ardekani B A, Ashburner J, Avants B, Chiang M C, Christensen G E, Collins D L, Gee J, Hellier P, Song J H, Jenkinson M, Lepage C, Rueckert D, Thompson P, Vercauteren T, Woods R P, Mann J J, Parsey R V. Evaluation of 14 nonlinear deformation algorithms applied to human brain MRI registration. Neuroimage, 2009, 46(3): 786-802 [19] Zhang Shao-Min, Zhi Li-Jia, Zhao Da-Zhe, Lin Shu-Kuan, Zhao Hong. Entropic graph estimation integrated with SIFT features for medical image non-rigid registration. Journal of Image and Graphics, 2012, 17(3): 412-418 (张少敏, 支力佳, 赵大哲, 林树宽, 赵宏. 融合SIFT特征的熵图估计医学图像非刚性配准. 中国图象图形学报, 2012, 17(3): 412- 418) [20] Liu C, Yuen J, Torralba A. Sift flow: dense correspondence across scenes and its applications. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(5): 978 -994
点击查看大图
计量
- 文章访问数: 1933
- HTML全文浏览量: 89
- PDF下载量: 1060
- 被引次数: 0