Spatial Video Watermarking Based on Stability of DC Coefficients

• Published in: Advances in Machine Learning and Cybernetics pp. 1033 - 1042
• Main Authors: Chen, Tianhang, Liu, Shaohui, Yao, Hongxun, Gao, Wen
• Format: Book Chapter Conference Proceeding
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg 2006; Springer
• Series: Lecture Notes in Computer Science
• Subjects: Applied sciences; Artificial intelligence; Computer science; control theory; systems; Exact sciences and technology; Memory and file management (including protection and security); Memory organisation. Data processing; Pattern recognition. Digital image processing. Computational geometry; Software; Blind; High performance; Correlation; Statistical analysis; Probabilistic approach; Video signal; Metadata; Video technique; Stéganography; Image compression; Cosine transform; Luminance; Direct current; Metric; Peak; Robustness; Safety; Digital rights management; High frequency; Discrete cosine transforms; key frame; Watermark; Artificial intelligence; Digital watermarking
• ISBN: 3540335846; 9783540335849
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/11739685_108

Watermarking technology plays an important role in digital rights management. Many high performance watermark algorithms have been developed. However, most of them focused on image media. This paper presents a novel blind video watermarking algorithm based on the stability of the direct current (DC) coefficient values in blocks. This algorithm first selects some frames by a key, and then embeds watermark into its luminance component of each selected frame. The luminance component to be watermarked first is divided into sub-blocks with size 8 by 8, and then discrete cosine transform (DCT) is used to process each sub-block to get the DC coefficients. Rearranging these DC coefficients into sub-sequences with equal length, and each such sub-sequence is transformed by DCT and get the highest frequency coefficient. Finally, the watermark is embedded in the high frequency coefficient array. The correlation metric is used in the detection process. The experimental results indicate that the correlation curve peaks emerge at the watermarked I frames and P frames which suffered from MPEG-4 and H.264 compression. The proposed watermarking method has strong robustness against some attacks such as frame deleting, frame inserting, frame shifting, frame statistical average and collusion attack.