Research On Robust Image And Video Watermarking Algorithms Against Camcorder Recording

With the development of the mobile Internet,the digital economy has produced an increasingly important impact on national life.Multimedia technology has playe d an important role in the Internet+ industry in recent years.Today's entertainment and education companies such as Toutiao,We Chat,and i Qiyi have massive media resources.These data themselves contain great value,and data protection has received more and more attention.As one of the important means of copyright protection for digital images,videos and other digital resources,watermarking technology can hide watermark information in the redundant data of images without affecting the value of the image itself,so as to achieve the purpose of copyright protection.Stimulated by the demand for applications,it has once again attracted the attention of the industry in recent years.After the watermark is embedded in the carrier,it often suffers from geome tric distortion,signal processing,spatial and temporal desynchronization,and camera recording along with the carrier,making the watermark ineffective.In this paper,the robust watermarking technology for digital images and video is studied,focusing o n the above-mentioned various attack methods,especially those that are closely related to actual applications.In this paper,an histogram image watermarking algorithm based on low-frequency construction is designed and extended to the video watermarking.At the same time,the characteristics of anti-camera recording attacks are considered,and for video copyright protection,a spatial-temporal context adaptive video watermarking algorithm is designed.In video watermarking,based on the observation that the time domain of adjacent frames is continuous,this paper embeds watermark information by customizing the luminance and chroma energy trends of the three channels of a YUV representation of consecutive frames.That is,this paper modulates the energy of consecutive frames into an ascending sequence to represent the watermark "-1" and a descending sequence to represent the watermark "1".In order to improve the robustness of watermark,this paper proposes a novel adaptive video watermarking scheme,which can adaptively embed and extract watermarks.In terms of embedding,the algorithm can automatically select the embedding area and embedding intensity according to the content characteristics of the video,and ensure the invisibility and robustness of the watermark.In terms of extraction,the algorithm can automatically determine the time domain embedding location of watermark by using the correlation relationship of watermark extracted from the luma and chroma channels to avoid time dessynchronization attack.At the same time,the extraction area is adaptively selected by the image matching method,which avoids the influence of the excessive difference between adjacent frames on the extraction and greatly improves the e xtraction accuracy.Experimental results show that the proposed method achieves high video quality and is robust to geometric attacks,compression,scaling,transcoding,recoding,frame rate conversion and camcorder recording.In the image watermarking algorithm,the watermark is embedded by modifying the relationship of the adjacent bars of the histogram.In the modified pixels,the lowfrequency characteristic of the image is used to assign a low-frequency weight to each pixel that may be modified.The pixel with the largest weight is modified to construct the low-frequency characteristic of the image,which makes the algorithm more robust to compression attacks.At the same time,it makes the modified image more smooth and improves the visual quality of the image.Finally,the watermarking algorithm is applied to the video using a spread spectrum strategy.So that the algorithm can be applied to the image and video,which improves the applicability of the algorithm.