A new video watermarking algorithm based on shot segmentation and block classification

In this paper, we propose a new video watermarking algorithm based on shot segmentation and block classification to enhance the robustness, imperceptibility and real-time performance based on the H.264/AVC codec. A method of selecting host frames is proposed based on shot segmentation to avoid embedding watermark frame by frame, so as to improve the robustness and the real-time performance. The watermark signal is cropped into small watermarks according to the number of shots in the host video, and small watermarks are respectively embedded into different shots. The watermarking capacity and the perceptual quality are greatly improved by this way. A method of selecting host coefficients is proposed based on block classification in the Discrete Cosine Transformation (DCT) compressed domain. The texture characteristics of host blocks are considered in the classification and the places of host coefficients can change adaptively according to the content of the video. The imperceptibility of the watermarked video is greatly improved by this way. The simplified quantization index modulation (QIM) is applied to embed watermark. It brings fewer artifacts to the host signal than the current main watermarking method, such as spread spectrum (SS), differential energy watermarking (DEW) and so on. The experiment results show that the proposed scheme has a good performance in maintaining real-time performance and resisting Gaussian noising, frame swapping, MPEG compression, etc.     

MCDCT-TF based video watermarking resilient to temporal and quality scaling

Heterogeneity among the end using display devices and corresponding varying system requirements make scalable version of H.264/AVC standard more popular. Existing watermarking based authentication system may not be suitable well for this new extension, because the scalability property of the codec may itself is perceived as an attack popularly known as content adaptation attack. In this paper, a blind scalable video watermarking scheme is proposed, which is robust against quality and temporal scalability. In the proposed scheme, Discrete Cosine Transform (DCT) based temporal filtering and wavelet based spatial filtering is used for choosing suitable watermark embedding zone. Experimental evidences are provided to justify the improved robustness of the proposed scheme over existing related watermarking schemes. The visual quality of the watermarked video has also been evaluated to show the efficiency of the proposed scheme.     

Scalable video transmission for the Internet

Information systems and distributed applications for the Internet show a growing demand for real-time audiovisual services. Support for those services within the Internet protocol layer will be widely available in the near future. Nevertheless, the heterogeneous structure of the Internet remains to be a great obstacle for establishing real-time video services. Scalable video codecs, generating bit-streams decodable at different rates, have been proposed to address the heterogeneity problem. In this article we present a scalable codec based on a spatio-temporal resolution pyramid combined with lattice vector quantization for efficient compression. With this approach additional lower bit-rate layers can be transmitted at the same overall bit-rate as needed by an H.263 single-layer transmission. The complexity of the codec is sufficiently low to allow software-only implementations of Internet video services. This is demonstrated on an Internet video server application.     

Content-based scalable H.263 video coding for road traffic monitoring

For sending video data through very low bit-rate mobile channels, video codec with high compression rate is the pre-requisite. Although the H.263 video codec is recommended as one of the candidates due to its simplicity and efficiency, it is generally believed that its compression efficiency can be further improved if the content-based scalable video coding technique can be applied. In this paper, we propose a modified H.263 encoder which supports real-time content-based scalable video coding. The proposed technique is applied to real-time video surveillance systems for road traffic monitoring. For the proposed approach, the moving objects, i.e. cars, are first extracted from the steady background. Their activities are then further classified as fast or slow by assessing the regularity of their motion. The information is then passed to a modified H.263 encoder to reduce the temporal and spatial redundancies in the video. As compared with the conventional H.263 encoder using for the same application, the proposed system has a 20% increase in compression rate with negligible visual distortion. The proposed system fully complies with the ITU H.263 standard hence the encoded bit stream is completely comprehensible to the conventional H.263 decoder.     

基于ARM11的H.264硬编解码视频传输系统设计

为了解决视频信息冗余、视频传输系统成本较高,视频编解码效率低等缺陷,研究了一种基于H.264硬编解码器的视频传输系统;该系统采用以ARM11为核心的,包含多媒体硬编解码器MFC的S3C6410作为处理器,使用CMOS摄像头采集实时视频数据;在嵌入式Linux操作系统上构建基于live555的RTSP服务器,监听客户端请求;重点阐述了多媒体硬编解码器MFC对视频数据进行的H.264硬件压缩编码,以及数据包的RTP封装,然后经由live555流媒体服务器转发至PC机;最后,在PC上对接收到的数据流进行解码播放,经实验测试证明,系统设计稳定可靠,具有可扩展性强、性能高、成本和功耗低等优点,图像质量和时延满足实际应用需求。     

Real-time video watermarking system on the compressed domain for high-definition video contents: Practical issues

Everyday, we encounter high-quality multimedia contents from HDTV broadcasting, DVD, and high-speed Internet services. These contents are, unhappily, processed and distributed without protection. This paper proposes a practical video watermarking technique on the compressed domain that is real-time and robust against video processing attacks. In particular, we focus on video processing that is commonly used in practice such as downscaling resolution, framerate changing, and transcoding. Most previous watermarking algorithms are unable to survive when these processings are strong or composite. We extract low frequency coefficients of frames in fast by partly decoding videos and apply a quantization index modulation scheme to embed and detect the watermark. On an Intel architecture computer, we implement a prototype system and measure performance against video processing attacks frequently occur in the real world. Simulation results show that our video watermarking system satisfies real-time requirements and is robust to protect the copyright of HD video contents.     

A real-time video watermarking system with buffer sharing for video-on-demand service

Real-time watermarking for streaming video (such as VOD service) requires significant amounts of computing resources. To address this issue, we present a scalable watermarking scheme integrated in a parallel MPEG-2 engine. A content-based block selection algorithm is proposed to efficiently embed the pseudo-random watermark signatures into DCT blocks. Our watermark scheme also provides a robust way to synchronize the watermarked video to the original source at detectors and is very resilient against cumulative and temporal attack.We optimize the parallel watermark engine to achieve real-time watermarking performance. We found that the system throughput could suffer significant degradation when processing high-level MPEG-2 video (such as HDTV) due to inefficient management of memory space. Therefore, we investigated an efficient buffer management scheme consisting of two methods: First we reduced the transmission buffer in slave nodes by frames sharing between frames in the Group-of-Picture (GOP) level. Then we further reduce the buffer space by a dynamic on-demand allocation on the slave side. By solving the memory-shortage bottleneck, the proposed system can support real-time watermarking for multiple high-resolution (up to 1404 × 960) video.     

基于WebRTC的实时视音频通信研究综述

实时网页通信(WebRTC)无需在网页上安装插件或客户端即可实现实时视音频通信,具有跨平台、开发成本低和适用范围广等优点,已被W3C纳为草案,正在带来多媒体通信的一场革命.首先介绍基于WebRTC实时视音频通信的总体架构;接着详细阐述其相关关键技术,包括信令机制、网页应用和浏览器底层WebRTC,后者还包括视、音频编译码模块和传输模块;然后从两用户和多用户的角度对WebRTC实时视音频通信应用的实现细节进行比较;最后讨论值得进一步研究的开放问题.     

基于整数小波变换和SVD的视频水印算法

提出了一种以二值图像为水印的混合整数小波变换和奇异值分解的视频水印盲提取算法。对水印图像进行混沌加密和Arnold置乱处理,选择计算复杂度低的直方图算法将视频分割为若干场景;借助密钥随机选取某些场景的亮度分量进行l级整数小波变换,再对低频子带进行分块的奇异值分解;采用量化的方法,将预处理后的水印图像嵌入奇异值分解后的最大奇异值中。在嵌入了水印的视频场景中提取所有的水印版本之后,利用对提取的所有水印信号版本进行统计求和的方法得到最终提取的水印图像。实验表明,提出的算法具有较好的透明性,对常见的处理具有较好的鲁棒性。     

Novel robust image watermarking based on subsampling and DWT

This paper presents a robust digital image watermarking scheme based on subsampling and DWT. Subsampling is firstly used to construct a subimage sequence as a video segment. Then, a random watermark sequence satisfied with Gaussian distribution is block-wised embedded into the DWT domain of these subimages repeatedly using the video watermarking technique. And watermark is detected through computing correlation between watermark and watermarked frames. The experiment results demonstrate that the proposed scheme achieves good robustness against JPEG compression, common image processing operation and geometric distortions. Furthermore, the proposed watermarking scheme is also robust against linear collusion and other video watermarking attacks.     

人眼视觉特性与SIFT相结合的视频双水印算法

针对视频水印容易遭受几何攻击以及水印的鲁棒性与透明性的平衡问题,提出一种基于人眼视觉特性与尺度不变特征变换（SIFT）相结合的抗几何攻击视频双水印算法。首先获取视频序列中人眼视觉掩蔽阈值作为水印的最大嵌入强度。其次,将视频帧进行离散小波变换（DWT）,对中高频子带系数提出基于视频运动信息的自适应水印算法;针对低频子带,提出基于小波低频系数统计特性的抗几何攻击视频水印算法。最后,以SIFT作为触发器判断视频帧是否遭受几何攻击,对遭受几何攻击的视频帧利用SIFT的尺度与方向不变性进行校正,并对校正后的视频帧提取水印信号;针对非几何攻击的视频帧,直接利用中高频提取算法。所提算法与实时性视频水印算法——基于小波域直方图的视频水印（VW-HDWT）算法比较,峰值性噪比值（PSNR）提高了7.5%;与基于特征区域的水印算法相比,水印嵌入容量提高约10倍。实验结果表明,在保证水印透明度较好的情况下,所提算法对常规几何攻击具有较强的鲁棒性。     

位域实时视频水印算法

在目前数字水印算法中，图像水印的算法远多于视频水印，然而在现实生活中，视频产品保护更加重要．视频水印除了满足图像水印的一般特性之外，还必须满足实时特性．因此视频水印的嵌入和提取算法必须有较低的计算复杂性．本文提出一种用于视频产品完整性认证的位域实时视频水印的算法．实验表明，该算法有较高的视觉质量，同时能对修改进行准确定位．满足完整性认证水印要求．     

鲁棒视频水印研究进展

数字视频在当前通信世界中被认为是一种重要而有效的媒体,广泛应用于新闻、短视频和有线网络广播视频节目中。随着计算机与互联网技术的发展,数字视频内容容易被侵权使用者肆意复制和传播,如何保护视频版权日益成为人们关注的问题。鲁棒视频水印是实现视频版权保护的一种有效手段。作为数字视频水印的分支,鲁棒视频水印是一种通过特定算法在需要被保护的视频对象中嵌入秘密信息——水印来证明版权归属的技术。本文对当前的视频水印技术进行了概述,对视频水印的概念、应用场景、分类方式、设计要求、发展历程和相关经典方法进行了介绍和梳理。本文归纳总结了2016—2021年鲁棒视频水印相关研究工作,包括基于内容的、基于码流的、基于深度学习和其他类型视频水印,并对其中部分工作进行了相应的性能比较和分析。其中,基于内容的视频水印方法将视频看做帧序列,由于在每一帧上应用水印算法,不考虑视频的编解码过程,这类方法实现简单,计算效率高;基于码流的视频水印方法将水印嵌入到编码比特流中,该方案更快速,故可支持实时视频水印应用;基于深度学习的方法取代了依靠手工设计的特征来提高水印的性能。最后分析了鲁棒视频水印的未来发展趋势。     

一种抗重压缩的H.264/AVC视频水印算法*

针对当前视频水印抗重压缩能力不足,提出一种基于I帧色度量化残差系数的视频水印算法。通过对H.264/AVC帧内预测模式的分析,得出重压缩后预测模式的转变是造成水印提取失败的主要因素,因此,算法选择在更加稳定的色度分量中嵌入水印。首先,利用4?4子块中非零系数个数选择合适的色度子块,然后,运用差分能量的思想,结合Arnold置乱后的水印信息,通过调制其中一个非零量化DCT残差系数改变两个特定位置系数的关系嵌入水印。算法在帧内预测模式色度分量中嵌入水印,有效的解决了重压缩后水印提取不同步问题,提高了水印的抗重压缩性能。实验结果显示,提取的水印图像位错误率(BER)在20%以下,重构图像的SSIM比现有方案提高了1.2%-4.4%,另外,测试序列的比特增加率(BIR)平均降低了70%左右。算法适用于广播监控等实时视频应用。     

无人机多跳自组网实时视频传输性能评估

基于当前无人机广泛采用的M-JPEG与MPEG-2视频编解码器,构建了实时半实物仿真环境,并基于此对无人机多跳自组网实时视频传输进行了性能评估。结果表明,实时条件下的M-JPEG编解码器无法支持多跳实时视频传输,而MPEG-2能够支持2跳至3跳的实时视频传输;在1~10跳的实时传输条件下,MPEG-2较之M-JPEG在分组投送成功率和解码率意义下的视频质量平均分别提高了55.69%和205.64%。相对于传统的单飞单控方式,无人机多跳自组网在将无人机系统的作用范围扩大4~9倍的同时能够支持实时视频传输。     

基于自动镜头分割的视频压缩编码算法

将镜头分割引入压缩编码,针对分割镜头特性确定适当的编码器及参数,能够提高视频压缩编码的效率．为此,提出了一种基于对分查找的镜头分割算法(VSSB)以及关键帧频率计算公式．其算法的核心是通过对分查找形成二叉树,同时通过广度优先搜索寻找分割点,具有O(log n)的时间复杂性．针对实时应用,采用了双阈值比较的镜头分割算法,能够较好地实现镜头切变和渐变的检测．基于镜头分割算法,建立并实现了并行方式的编解码系统模型．实验证明,基于镜头分割的编码方法能够提高压缩比,缩短压缩编码时间,有利于改善重建图像质量。     

基于SDL的Mpeg-4视频流实时解码与回放方法

主要研究一个基于跨平台多媒体软件开发库——SDL的视频流实时解码与回放实现方法。主要利用SDL对线程的控制来实现实时,SDL调用视频解码器来实现解码工作,利用SDL本身对YUV格式的视频数据处理来实现回放。在IGHz处理器的PC上的测试表明,可获得20帧／秒的解码播放效果,并且画面流畅,而且程序具有良好的跨平台性。     

数据驱动的并行视频编码器的设计和实现(英文)

通用视频编码系统一般基于标准视频压缩算法, 比如 MPEG4 和 H.264。但这些标准算法包括如离散余弦变换和运动估计等高复杂度的计算,所以为了实现实时通信,编码系统通常采用以硬件为基础的实现方法。然而以硬件为基础的视频编码系统具有实施周期长、成本高、灵活性差等不足。文章介绍了一种基于低计算复杂度自适应块截短编码方案的数据驱动并行视频编码系统,实验结果验证了编码系统的有效性。     

基于纹理方向的高效视频编码鲁棒视频水印

针对现有基于高效视频编码的视频水印算法鲁棒性不足的问题,提出一种基于纹理方向的鲁棒视频水印算法。算法根据水印值将帧内角度预测模式分为水平方向和垂直方向两组,计算N×N分割编码时每个预测单元的纹理方向。当纹理方向与水印所表示的方向一致时,将当前预测单元的33种角度预测模式截断为水平方向或者垂直方向的预测模式,随后根据率失真代价函数决策出最优的预测模式,并判断水印是否嵌入。记录最终水印嵌入位置作为密钥,用于解码端提取。实验结果表明,该算法具有较小的码率增长和视频失真,在历经噪声、滤波和重编码等攻击后能保持较低的误码率(BER),说明该方案能用于视频版权保护。     

基于深度学习的实时场景小脸检测方法

实时场景下的小脸检测存在检出率低而且回归精度差的问题。通过融合更底层特征进行多尺度级联预测。根据实时场景下的人脸特点生成不同大小和比例的预测框以更好地适应人脸形状。在预测阶段提出了基于IOU判别的soft and hard nms算法，对冗余预测框进行抑制，设置两个阈值将网络生成的预测框划分为低中高三段，对不同段的预测框采取不同的处理以达到精准筛选的目的。最优架构可在两张NVIDIA GTX 1080显卡下的实时视频检测和摄像头检测中获得45 f/s的速度，并且在Wider Face总体验证集上取得82.6%的平均精度。