TY - JOUR
T1 - Universal joint image clustering and registration using multivariate information measures
AU - Raman, Ravi Kiran
AU - Varshney, Lav R.
N1 - Funding Information:
Manuscript received November 30, 2017; revised May 2, 2018; accepted June 26, 2018. Date of publication July 11, 2018; date of current version September 27, 2018. This work was supported by the Air Force STTR under Grant FA8650-16-M-1819. This paper was presented in part at the 2017 IEEE International Symposium on Information Theory, Aachen, Germany, Jun 25–30, 2017. The guest editor coordinating the review of this paper and approving it for publication was Dr. Stark Draper. (Corresponding author: Ravi Kiran Raman.) The authors are with the Coordinated Science Laboratory and the Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Champaign, IL 61801 USA (e-mail:, rraman10@ illinois.edu; varshney@illinois.edu).
Publisher Copyright:
© 2007-2012 IEEE.
PY - 2018/10
Y1 - 2018/10
N2 - We consider the problem of universal joint clustering and registration of images. Image clustering focuses on grouping similar images, while image registration refers to the task of aligning copies of an image that have been subject to rigid-body transformations, such as rotations and translations. We first study registering two images using maximum mutual information and prove its asymptotic optimality. We then show the shortcomings of pairwise registration in multi-image registration, and design an asymptotically optimal algorithm based on multi-information. Further, we define a novel multivariate information functional to perform joint clustering and registration of images, and prove consistency of the algorithm. Finally, we consider registration and clustering of numerous limited-resolution images, defining algorithms that are order-optimal in scaling of number of pixels in each image with the number of images.
AB - We consider the problem of universal joint clustering and registration of images. Image clustering focuses on grouping similar images, while image registration refers to the task of aligning copies of an image that have been subject to rigid-body transformations, such as rotations and translations. We first study registering two images using maximum mutual information and prove its asymptotic optimality. We then show the shortcomings of pairwise registration in multi-image registration, and design an asymptotically optimal algorithm based on multi-information. Further, we define a novel multivariate information functional to perform joint clustering and registration of images, and prove consistency of the algorithm. Finally, we consider registration and clustering of numerous limited-resolution images, defining algorithms that are order-optimal in scaling of number of pixels in each image with the number of images.
KW - Image registration
KW - asymptotic optimality
KW - clustering
KW - universal information theory
KW - unsupervised learning
UR - http://www.scopus.com/inward/record.url?scp=85049797741&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85049797741&partnerID=8YFLogxK
U2 - 10.1109/JSTSP.2018.2855057
DO - 10.1109/JSTSP.2018.2855057
M3 - Article
AN - SCOPUS:85049797741
SN - 1932-4553
VL - 12
SP - 928
EP - 943
JO - IEEE Journal on Selected Topics in Signal Processing
JF - IEEE Journal on Selected Topics in Signal Processing
IS - 5
M1 - 8409946
ER -