A secure image authentication algorithm with pixel-level tamper localization

• Published in: 2004 International Conference on Image Processing, 2004. ICIP '04 Vol. 3; pp. 1573 - 1576 Vol. 3
• Main Authors: Jinhai Wu, Zhu, B.B., Shipeng Li, Fuzong Lin
• Format: Conference Proceeding
• Language: English
• Published: Piscataway NJ IEEE 2004
• Subjects: Applied sciences; Artificial intelligence; Asia; Authentication; Computer science; control theory; systems; Exact sciences and technology; Gray-scale; Intelligent systems; Pattern recognition. Digital image processing. Computational geometry; Pixel; Signal design; Watermarking; Safety; Image processing; Localization; Oracle; Modeling; Authentication
• ISBN: 0780385543; 9780780385542
• ISSN: 1522-4880
• DOI: 10.1109/ICIP.2004.1421367

In this paper, we use a generalized model for all the previously proposed pixel-wise authentication schemes. Then we show how this model can be compromised with an Oracle attack. This motivated us to develop a novel authentication scheme to be described here. The new scheme is designed to detect any changes to a signal. It consists of two mechanisms dedicated to authenticity verification for a signal and localization of tampered pixels, respectively. The scheme is secure yet it maintains the fine tamper localization capability of a pixel-wise scheme. Experimental results show that the proposed scheme has a good accuracy in locating tampered pixels.