Multimodal Retinal Image Registration method Based on Speed-up Robust Feature

DOI

10.2991/wartia-16.2016.137How to use a DOI?

Keywords

Multimodal retinal image, Registration, Feature matching, Transformation model, Random sample consensus

Abstract

Retinal fundus image plays an important role in the diagnosis of eye diseases. In order to have a comprehensive understanding of retinal image, doctors need to take pictures from different angles of the patient's eyes. Because of the movement of the eye and head of the patient, there are some differences between the corresponding physiological structures. So it is necessary to carry out registration before information fusion. In this paper, a registration method based on speed-up robust feature is proposed. Firstly, preprocessing is performed on the fundus images, Secondly, surf feature points are extracted from the reference image and the floating image, and the descriptors are generated later. Finally, the spatial transformation of the floating image is carried out. In this paper, 120 fundus images of 20 patients were tested, the results showed that for different resolutions, angles, and translations, our method can be achieved automatic registration, average matching rate can reach 93.3%, and precision can reach sub pixel level.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Zhitao Xiao
AU  - Wan Zhu
AU  - Fang Zhang
AU  - Jun Wu
AU  - Lei Geng
AU  - Wen Wang
PY  - 2016/05
DA  - 2016/05
TI  - Multimodal Retinal Image Registration method Based on Speed-up Robust Feature
BT  - Proceedings of the 2016 2nd Workshop on Advanced Research and Technology in Industry Applications
PB  - Atlantis Press
SP  - 656
EP  - 662
SN  - 2352-5401
UR  - https://doi.org/10.2991/wartia-16.2016.137
DO  - 10.2991/wartia-16.2016.137
ID  - Xiao2016/05
ER  -