Abstract
Compared with pairwise registration, the groupwise one is capable of handling a large-scale population of images simultaneously in an unbiased way. In this work we improve upon the state-of-the-art pixel-level, Least-Squares (LS)-based groupwise image registration methods. Specifically, the registration technique is properly adapted by the use of Self Quotient Images (SQI) in order to become capable for solving the groupwise registration of photometrically distorted, partially occluded as well as unimodal and multimodal images. Moreover, the proposed groupwise technique is linear to the cardinality of the image set and thus it can be used for the successful solution of the problem on large image sets with low complexity. From the application of the proposed technique on a series of experiments for the groupwise registration of photometrically and geometrically distorted, partially occluded faces as well as unimodal and multimodal magnetic resonance image sets and its comparison with the Lucas–Kanade Entropy (LKE) algorithm, the obtained results look very promising, in terms of alignment quality, using as figures of merit the mean Peak Signal to Noise Ratio () and mean Structural Similarity (), and computational cost.
Highlights
Groupwise image alignment/registration or congealign is a joint alignment process that handles a large scale of images simultaneously, in contrast to pairwise alignment/registration
Congealign algorithms tend to utilize one image at a time as the held out image and the rest as the stack, that keep changing while a dissimilarity/similarity function is iteratively minimized/maximized
This is done by estimating warp updates for each image that best align them with the stack
Summary
Groupwise image alignment/registration or congealign is a joint alignment process that handles a large scale of images simultaneously, in contrast to pairwise alignment/registration. Congealign algorithms tend to utilize one image at a time as the held out image and the rest as the stack, that keep changing while a dissimilarity/similarity function is iteratively minimized/maximized. This is done by estimating warp updates for each image that best align them with the stack. In [4] entropy-based congealign was introduced, by defining a method that minimizes the joint entropy across pixel stacks distributions. In [5], entropy-based congealing was adapted to handle large sets of gray-scale valued, 3D medical images. In [7] and [8] Least
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
