Image authentication scheme based on self-embedding watermarking

7Citations
Citations of this article
8Readers
Mendeley users who have this article in their library.

This article is free to access.

Abstract

This paper presents a block-wise and content-based semi-fragile image watermarking authentication scheme with location and recovery capability. Firstly the image is segmented by two regions: Region of Interest (ROI) and Region of Embedding (ROE). The watermark sequence is extracted from ROI and it is embedded into the middle frequency band of DCT coefficients of ROE. In the authentication stage, two watermark sequences extracted ROI and ROE, respectively, are used. If difference between both sequences of a ROI block is smaller than the predefined threshold value, the ROI block is determined authentic, otherwise the block is considered as tampered and it is recovered by the recovery process. The proposed scheme is evaluated from several points of view: watermark imperceptibility, capability of tamper detection, image quality of recovered regions and robustness of no-intentional manipulations, such as JPEG compression. The simulation results show fairly good performance of the proposed scheme. © 2009 Springer-Verlag Berlin Heidelberg.

Cite

CITATION STYLE

APA

Cruz-Ramos, C., Reyes-Reyes, R., Nakano-Miyatake, M., & Pérez-Meana, H. (2009). Image authentication scheme based on self-embedding watermarking. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5856 LNCS, pp. 1005–1012). https://doi.org/10.1007/978-3-642-10268-4_117

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free