Video Watermarking Combining with Hybrid Coding
Liang Jian-xun · Journal of Circuits and Systems · 2003
A watermarking scheme combining with hybrid video coding is proposed. The intra-coded frame, which is called intra-frame, is blocked and discrete cosine transformed (DCT), watermark is embedded in the low-frequency coefficients in DCT domain. While for the inter-coded frame, the blocked motion compensation is discrete cosine transformed as in I frame, then the DC coefficient in the DCT matrix is watermarked. In this scheme, human perceptual model is considered. In the process of embedding, spread spectrum technology, which is originally used in communication, is cooperated with multidimensional watermarking technique. The watermark detection is based on the correlation and hypothesis testing and directly put on the video stream before decoding. For these reasons, computational complex can be greatly reduced to increase the usability of video watermarking.