DocumentCode
185784
Title
Semi-fragile watermarking for image tamper localization and self-recovery
Author
Yuhang Li ; Ling Du
Author_Institution
Inf. Network Center, Shenyang Aerosp. Univ., Shenyang, China
fYear
2014
fDate
18-19 Oct. 2014
Firstpage
328
Lastpage
333
Abstract
This paper proposes a semi-fragile digital watermarking scheme for image tamper localization and self-recovery. Firstly, the authentication watermark is generated for tampered area localization. After that, for content recovery, the recovery watermark is calculated from high frequency band with redundancy free. Then, for the two type of watermarks, the DWT-based watermark embedding method makes our proposed scheme tolerable against friendly manipulations. Even some parts of the watermarked image are tampered and the corresponding embedded information are missing, we can still recover the image from the untouched area based on the localization results and compressive sensing. Moreover, the encryption algorithm based on chaotic function guarantees the safety of our two type of watermarks. Extensive experiments demonstrate the effectiveness of the proposed scheme, which is sensitive against malicious tamper and robust to some incidental manipulations (e.g. JPEG compression, slight noise addition, brightness/contract adjustment and format conversion).
Keywords
chaos; compressed sensing; discrete wavelet transforms; image coding; image watermarking; message authentication; DWT-based watermark embedding method; authentication watermark; chaotic function; compressive sensing; content recovery; encryption algorithm; image self-recovery; image tamper localization; image watermarking; semifragile digital watermarking scheme; tampered area localization; Authentication; Encryption; PSNR; Robustness; Vectors; Watermarking;
fLanguage
English
Publisher
ieee
Conference_Titel
Security, Pattern Analysis, and Cybernetics (SPAC), 2014 International Conference on
Conference_Location
Wuhan
Print_ISBN
978-1-4799-5352-3
Type
conf
DOI
10.1109/SPAC.2014.6982711
Filename
6982711
Link To Document