Paper Abstract and Keywords |
Presentation |
2018-05-17 15:15
On OCT Volumetric Data Restoration via Hierarchical Sparsity and Hard Constraint Shogo Muramatsu, Satoshi Nagayama, Samuel Choi (Niigata Univ.), Shunsuke Ono (Tokyo Institute of Tech.), Takeru Ota, Fumiaki Nin, Hiroshi Hibino (Niigata Univ.) SIP2018-3 IE2018-3 PRMU2018-3 MI2018-3 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
This work proposes a novel restoration method for optical coherence tomography (OCT) data. OCT is a measurement technique that acquires a tomographic image in the order of a few $mu$m scale by using a near infrared laser. In order to deeply understand the mechanism of sensory epithelium in cochlear, the authors are working on dynamic tomographic imaging of hair cells of guinea pig in vivo and develop a multi-frequency swept (MS) enface OCT device. Currently, the device faces to a problem that the reflected light is quite weak, and a help of image processing is demanded to estimate the reflectional distribution hidden in noise. In the article cite{icassp2018}, a restoration method is given by reducing the problem to a regularized least squares one with a hard constraint for reflections. This work further proposes to introduce another regularization and constraint for the latent refractive index distribution. The significance of the proposed model is verified by simulation on artificial data. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Volumetric data restoration / Primal-dual proximal splitting method / Sparse regularization / Hard constraint / BM4D / OCT / / |
Reference Info. |
IEICE Tech. Rep., vol. 118, no. 33, SIP2018-3, pp. 7-12, May 2018. |
Paper # |
SIP2018-3 |
Date of Issue |
2018-05-10 (SIP, IE, PRMU, MI) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
SIP2018-3 IE2018-3 PRMU2018-3 MI2018-3 |
|