Loughborough University
Leicestershire, UK
LE11 3TU
+44 (0)1509 263171
Loughborough University

Loughborough University Institutional Repository

Please use this identifier to cite or link to this item: https://dspace.lboro.ac.uk/2134/32538

Title: Discriminative and robust zero-watermarking scheme based on completed local binary pattern for authentication and copyright identification of medical images
Authors: Liu, Xiyao
Lou, Jieting
Wang, Yifan
Zou, Beiji
Chen, Yan
Keywords: Medical image
Copyright identification
Completed local binary pattern
Issue Date: 2018
Publisher: © Society of Photo-Optical Instrumentation Engineers (SPIE)
Citation: LIU, X. ... et al, 2018. Discriminative and robust zero-watermarking scheme based on completed local binary pattern for authentication and copyright identification of medical images. IN: Zhang, J. and Chen, P-H. (eds). Proc. SPIE 10579, Medical Imaging 2018: Imaging Informatics for Healthcare, Research, and Applications, 105791I (6 March 2018); doi: 10.1117/12.2292852.
Abstract: Authentication and copyright identification are two critical security issues for medical images. Although zerowatermarking schemes can provide durable, reliable and distortion-free protection for medical images, the existing zerowatermarking schemes for medical images still face two problems. On one hand, they rarely considered the distinguishability for medical images, which is critical because different medical images are sometimes similar to each other. On the other hand, their robustness against geometric attacks, such as cropping, rotation and flipping, is insufficient. In this study, a novel discriminative and robust zero-watermarking (DRZW) is proposed to address these two problems. In DRZW, content-based features of medical images are first extracted based on completed local binary pattern (CLBP) operator to ensure the distinguishability and robustness, especially against geometric attacks. Then, master shares and ownership shares are generated from the content-based features and watermark according to (2,2) visual cryptography. Finally, the ownership shares are stored for authentication and copyright identification. For queried medical images, their content-based features are extracted and master shares are generated. Their watermarks for authentication and copyright identification are recovered by stacking the generated master shares and stored ownership shares. 200 different medical images of 5 types are collected as the testing data and our experimental results demonstrate that DRZW ensures both the accuracy and reliability of authentication and copyright identification. When fixing the false positive rate to 1.00%, the average value of false negative rates by using DRZW is only 1.75% under 20 common attacks with different parameters.
Description: Copyright (2018) Society of Photo-Optical Instrumentation Engineers (SPIE). One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.
Sponsor: This research is supported by the National Natural Science Foundations of China (61602527, 61573380), Hunan Provincial Natural Science Foundation of China (2017JJ3416) and China Postdoctoral Science Foundation (2017M612585).
Version: Published
DOI: 10.1117/12.2292852
URI: https://dspace.lboro.ac.uk/2134/32538
Publisher Link: https://doi.org/10.1117/12.2292852
ISSN: 0277-786X
Appears in Collections:Conference Papers and Presentations (Computer Science)

Files associated with this item:

File Description SizeFormat
105791I.pdfPublished version516.88 kBAdobe PDFView/Open


SFX Query

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.