A RST Invariant Watermarking Scheme Utilizing Support Vector Machine and Image Moments for Synchronization

Author

Wu Jianzhen

Author_Institution

Dept. of Autom., Shanghai Univ. of Eng. Sci., Shanghai, China

Volume

1

fYear

2009

fDate

18-20 Aug. 2009

Firstpage

572

Lastpage

574

Abstract

A rotation, scaling and translation (RST) invariant watermarking scheme using support vector machine and image moments for synchronization is proposed in this paper. Watermark is adaptively embedded in discrete wavelet transform (DWT) domain. In order to estimate RST transform parameters, support vector machine is utilized to learn image geometric pattern represented by six combined low order image moments. Watermark extraction is carried out after watermarked image has been synchronized. Experiments show that the proposed scheme can effectively resist JPEG compression, noise and geometric attacks.

Keywords

discrete wavelet transforms; geometry; image coding; image representation; learning (artificial intelligence); support vector machines; watermarking; discrete wavelet transform; geometry; image moment; image representation; invariant watermarking scheme; support vector machine; Detectors; Discrete transforms; Discrete wavelet transforms; Image coding; Image registration; Information security; Parameter estimation; Support vector machines; Watermarking; Wavelet domain; gemotric attacks; image moment; image registration; robust; support vector machine;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Assurance and Security, 2009. IAS '09. Fifth International Conference on

Conference_Location

Xian

Print_ISBN

978-0-7695-3744-3

Type

conf

DOI

10.1109/IAS.2009.176

Filename

5283758