DeepAI AI Chat
Log In Sign Up

Deformable Groupwise Image Registration using Low-Rank and Sparse Decomposition

by   Roland Haase, et al.

Low-rank and sparse decompositions and robust PCA (RPCA) are highly successful techniques in image processing and have recently found use in groupwise image registration. In this paper, we investigate the drawbacks of the most common RPCA-dissimilarity metric in image registration and derive an improved version. In particular, this new metric models low-rank requirements through explicit constraints instead of penalties and thus avoids the pitfalls of the established metric. Equipped with total variation regularization, we present a theoretically justified multilevel scheme based on first-order primal-dual optimization to solve the resulting non-parametric registration problem. As confirmed by numerical experiments, our metric especially lends itself to data involving recurring changes in object appearance and potential sparse perturbations. We numerically compare its peformance to a number of related approaches.


page 4

page 13

page 16

page 17

page 18

page 20


Efficient Registration of Pathological Images: A Joint PCA/Image-Reconstruction Approach

Registration involving one or more images containing pathologies is chal...

Low-rank geometric mean metric learning

We propose a low-rank approach to learning a Mahalanobis metric from dat...

Sparse Online Low-Rank Projection and Outlier Rejection (SOLO) for 3-D Rigid-Body Motion Registration

Motivated by an emerging theory of robust low-rank matrix representation...

Sparse And Low Rank Decomposition Based Batch Image Alignment for Speckle Reduction of retinal OCT Images

Optical Coherence Tomography (OCT) is an emerging technique in the field...

A low-rank representation for unsupervised registration of medical images

Registration networks have shown great application potentials in medical...

A Novel Similarity Measure for Image Sequences

Quantification of image similarity is a common problem in image processi...

Unseeded low-rank graph matching by transform-based unsupervised point registration

The problem of learning a correspondence relationship between nodes of t...