Embedding and detection method and system of large-capacity video digital watermark

Abstract

The invention provides a method and a system for embedding and detecting large-capacity video digital watermarks. The method includes: disintegrating a video watermark W into a watermark head WH and M sub-watermarks WS (I); dividing all video frames F of original video into M+1 groups of video frames to obtain a group of video frames FH and M groups of video frames FS (I); embedding an internal template to the groups of video frames FH to obtain a group of video frames FHT; embedding the watermark head WH to the group of video frames FHT to obtain a group of video frames FHW; respectively embedding the M sub-watermarks WS (I) to the M groups of video frames FS (I) to obtain M groups of video frames FSW (I); combining the groups of video frames FHW and the M groups of video frames FSW (I) sequentially to obtain a watermark-embedded video signal FW. The large-capacity video digital watermarks are embedded to the video, so that attacks such as spatial-domain geometric attack, lossy compression, video frame deletion, video frame insertion, video frame equalization and video clipping can be resisted effectively.

Description

technical field [0001] The invention belongs to the technical field of video digital watermarking, and in particular relates to a large-capacity video digital watermarking embedding and detection method and system thereof. Background technique [0002] Digital watermarking is a technology that embeds and detects digital marks in digital images, digital videos, digital audio or digital document carriers, and can be widely used in digital content retrieval, advertising monitoring, copyright identification, piracy tracking and other application fields. [0003] In 1954, Emil Hembrooke proposed the patent "Identification of sound and like signals", announcing the birth of digital watermarking technology. This patent embeds an identification code into music in an imperceptible way, thereby realizing the identity authentication of music producers. At this time, digital watermarking did not become an independent subject. Until the early 1990s, digital watermarking gradually became ...

AI Technical Summary

Problems solved by technology

[0006] Watermark capacity, invisibility, and robustness are three important indicators in the field of video watermarking. The existing video digital watermarking technology mainly has the following four problems: First, the watermarking capacity that can be embedded in a single frame image is small, which is difficult to meet According to the requirements of the application, it is necessary to use multiple frames in the video to expand the watermark capacity; secondly, the video frames are usually played continuously, and the general watermarking technology only embeds the watermark in the I frame, and does not embed the watermark in the P frame and B frame, resulting in There is a huge difference in image distortion between adjacent frames, which is prone to flickering during video playback; in addition, video timeline distortions such as frame deletion, frame averaging, frame rate conversion, and lens cropping will cause adjacent frames to interfere with each other and destroy the video. The sequence between frames can easily cause errors in video watermark detection; finally, spatial domain geometric attacks such as translation, shearing, zooming, and rotation can easily destroy the spatial synchronization of watermarks, resulting in watermarks that cannot be detected normally

Images