This item is available under a Creative Commons License for non-commercial use only
1.2 COMPUTER AND INFORMATION SCIENCE
Watermarking is a major image processing application used to authenticate user documents by embedding and hiding some authenticated piece of information behind an image, audio or the video file. For example, copyright symbols or signatures are often used. Our proposed work is to develop and implement an improved layered approach to video watermarking. The traditional watermarking approach tends to embed an entire watermark image within each video frame or within random video frames to give the appearance of a hidden watermark to the casual observer. This work proposes a more efficient and secure approach to perform watermarking, by using sub image classification. That is to say, selected frames only will contain a fractional number of pixels from the watermark image. We take k bits from the watermark and store then within a video frame, depending on the size of that watermark image. Our algorithm is capable of hiding high capacity information over video frames. The novel approach is to partially distribute the watermarking data over a set of frames until the entire watermark is eventually distributed throughout the entire video. The originality our technique is that it is a histogram inspired and reversible watermarking approach as defined with visual cryptography. Our approach hides similar watermarking pixels with frames of a similar appearance. Differing sets of watermark pixels are thus embedded within dissimilar frames, thus making the system more robust. It will provide a high degree of authentication, as the extraction of information from a single frame only will not reveal the entire watermarking data, or even give any obvious indication that it contains a fraction of the watermark pixels. The resilience of our technique will be tested by performing various systematic attacks upon a series of videos watermarked in this manner.
Kumar, M. and Hensman, A. Robust Digital Video Watermarking using Reversible Data Hiding and Visual Cryptography. Presented in ISSC conference 2013 at LYIT, Ireland.