H.264 video watermark embedding and extraction method based on mixed coding/decoding

Abstract

The invention discloses an H.264 video watermark embedding and extraction method based on mixed coding / decoding and belongs to the field of multimedia information safety. According to the scheme, video data are not required to be thoroughly decoded, coding parameters of an original video are fully utilized during watermark embedding, and difference between a video with an embedded watermark and the original video is reduced to the greatest extent. During decoding, not all of the video data are decoded, and only discrete cosine transformation (DCT) data are decoded. During the watermark embedding, the coding parameters such as a block type, a prediction mode and a motion vector in a decoding process are kept, and the watermark embedding is quickly realized. Compared with a conventional video watermark scheme in which all of the video data are required to be decoded, a video watermark scheme based on a 264 video compression standard provided by the invention is higher in speed and higher in efficiency.

Description

Technical field: [0001] The invention is a video watermark embedding and extraction scheme based on the H.264 video format of mixed coding and decoding, aims at embedding watermarks in DCT coefficients in the H.264 code stream, and belongs to the field of multimedia information security. Background technique: [0002] Digital watermarking is a very effective digital copyright protection technology, which has developed rapidly in recent years. It uses the visual characteristics of the human eye to embed symbol information that does not affect the original carrier's visual effect and use value in the digital carrier, and can extract the embedded watermark information through a special method during detection. However, the current research on digital watermarking is mostly concentrated in the field of static images. Some existing image watermarking algorithms can be applied to video watermarking algorithms, but video watermarking is not a simple extension of image watermarking...

AI Technical Summary

Problems solved by technology

Corresponding to these three embedding positions, there are three watermark embedding schemes: 1. Embedding directly into the original video data stream, the advantage is that there are many watermark embedding methods, and image watermark embedding methods can be used, but the disadvantage is that the calculation amount is large, which increases The video code stream reduces the quality of the video; correspondingly, when extracting the watermark, it is also necessary to decode the code stream and then extract the original video, which increases the complexity of the algorithm

2. The watermark is embedded in the video compression coding process. The common method is to embed the watermark into the quantization coefficient or motion vector. The advantage is that the embedded watermark information can be well combined with the video and will not increase the video stream. Data bit rate, the disadvantage is that it will reduce the video quality, and the watermark embedding algorithm should be selected in combination with the encoder

3. The watermark is directly embedded in the compressed bit stream. This scheme has the least impact on the video signal, and does not need to be encoded and decoded. It has high real-time performance, but it is limited by the compressed bit rate. This method limits the data embedded in the watermark. amount of size

If the watermark is directly embedded in the original video, many image watermark algorithms can be directly transplanted to the video watermark, but the disadvantage is that it will increase the video stream and reduce the quality of the video; Compression standards (such as the common MPEG-2 or MPEG-4) to embed watermarks in compressed video

Images