Title :
A wavelet-based semi-fragile watermarking with recovery mechanism
Author :
Tsai, MinJen ; Chien, ChihCheng
Author_Institution :
Inst. of Inf. Manage., Nat. Chiao Tung Univ., Hsinchu
Abstract :
This paper studied the wavelet transformation based watermarking tamper detection and recovery scheme. By the property of discrete wavelet transform, we design the semi-fragile watermark from low-frequency band and embedding the watermark into the high-frequency band by the help of human visual system (HVS). The watermarked image still preserves high image quality after the embedding of the semi-fragile watermark and the image authentication system is able to locate precisely any malicious alteration made to the image. Our novel design can restore the altered or destroyed regions from the attacked images and allow the users to know what the original content was in the manipulated areas. We also analyze the robustness for mild modification like JPEG compression and channel additive white Gaussian noise (AWGN) and fragility to malicious attack. In summary, this scheme can resist the mild modifications of digital image and be able to detect the malicious modifications precisely.
Keywords :
AWGN channels; data compression; discrete wavelet transforms; image coding; message authentication; watermarking; AWGN; JPEG compression; channel additive white Gaussian noise; digital image; discrete wavelet transform; human visual system; image authentication system; image quality; malicious attack; semifragile watermarking; watermarking tamper detection; watermarking tamper recovery; AWGN; Authentication; Discrete wavelet transforms; Humans; Image quality; Image restoration; Noise robustness; Transform coding; Visual system; Watermarking; semi-fragile; tamper detection; tamper recovery copyright protection; wavelet;
Conference_Titel :
Circuits and Systems, 2008. ISCAS 2008. IEEE International Symposium on
Conference_Location :
Seattle, WA
Print_ISBN :
978-1-4244-1683-7
Electronic_ISBN :
978-1-4244-1684-4
DOI :
10.1109/ISCAS.2008.4542097