压缩域快速视频拷贝检测算法

为了在大规模数据下有效提升拷贝检测速度,提出了一种压缩域视频拷贝检测算法,采用2级匹配框架,利用压缩域上DCT系数顺序度量特征进行相似度匹配,显著减少了查询视频解码操作;并为DCT系数顺序度量特征建立倒排索引,以加快第一级匹配的速度.实验结果表明,与已有的算法相比,所提出算法在保持相当的检测精度条件的同时,大幅提升了检测速度.     

基于数字签名的高效视频拷贝检测算法研究

为了解决视频特征提取和大型数据库的快速搜索问题,提出了一种新的视频拷贝检测算法。该算法基于视频剪辑的时空特征计算视频签名,通过视频签名的距离来衡量视频之间的相似性。为视频签名数据库建立聚类索引表,提高搜索速度。在大型数据库中查询测试的实验结果表明该算法对视频拷贝检测有较高查全率和正确率。该算法可以作为一种有效的视频跟踪和检测技术应用于数字广播电视的监控领域。     

改进模型估计的无人机侦察视频快速拼接方法

针对提高无人机侦察视频的拼接速度与效果的问题,提出一种改进模型估计的无人机侦察视频快速拼接方法。首先,基于自适应鲁棒性尺度不变的特征检测子对视频各帧进行基于点的特征匹配。其次,提出改进的随机抽样一致性算法进行模型估计,并去除误匹配点。最后,提出侦察影像快速拼接算法,计算各影像变换到正射拼接图的单应性矩阵,完成视频序列拼接。实验结果表明:改进的随机抽样一致性算法在保证鲁棒性的同时,提高了执行速度;侦察影像快速拼接算法提高了拼接速度,同时改善了拼接效果。     

基于多特征融合的视频检索算法

随着视频等多媒体数据呈指数式迅猛增长,高效快速的视频检索算法引起越来越多的重视。传统的图像特征如颜色直方图以及尺度不变特征变换等对视频拷贝检测中检索速度以及检测精度等问题无法达到很好的效果,因此文中提出一种多特征融合的视频检索方法。该方法利用前后两帧的时空特征进行基于滑动窗口的时间对齐算法,以达到减少检索的范围和提高检索速度的目的。该算法对关键帧进行灰度序列特征、颜色相关图特征以及SIFT局部特征提取,然后融合全局特征和局部特征两者的优势,从而提高检测精度。实验结果表明,该方法可达到较好的视频检索精度。     

基于Wu-Manber算法的大规模URL模式串匹配算法

大规模高速URL匹配是许多网络安全系统中的关键技术,经典串匹配算法在大规模URL情况下有许多限制.针对URL数据的特点在经典多模式串匹配算法Wu-Manber基础上提出XWM-Tree算法和XWM-Hash算法.算法应用了模式串窗口选择,两阶段哈希和关联容器组织冲突链表等多种优化手段,大幅度提高了算法的匹配性能.在大规模真实数据集上的测试结果表明本文提出的算法匹配速度可以提高一倍以上,尤其是当最短模式串较长的时候更有优势.     

基于CSS角点检测的快速匹配算法

为了实现准确、快速的图像匹配,从角点检测与描述子两方面入手,提出了一种基于CSS角点检测的匹配算法。首先,在曲率尺度空间下,检测图像在不同尺度下的角点并剔除不稳定角点;其次,基于曲率对图像轮廓描述的精确性,以特征点为中心划分3×4的子邻域,计算子邻域内轮廓曲线点的高斯加权曲率等4维向量特征,建立48维描述子,由于CSS角点检测包含曲率计算,因此生成描述子时避免了曲率的二次计算,提高了匹配速度;最后,提出一种"二进制距离"方法对描述子进行匹配,进一步优化匹配速度。通过实验证明,在保证精度的情况下,CSS快速匹配算法大幅度缩短了匹配时间,对旋转、亮度变化具有较好的匹配效果。     

GPU解码数据快速拷贝的应用研究

针对高清视频在客户端解码播放过程中存在的CPU占用率高、图像数据拷贝速度低等问题,提出了一种基于GPU解码数据快速拷贝方法。研究了DXVA硬解码方法在视频解码运算过程中的应用,为了消除解码数据在显存拷贝时产生的CPU占用率高现象,利用显存特点和SSE41多媒体指令新特性,设计并实现了视频帧数据快速拷贝方案。实验结果表明,该方法能在满足高清视频实时播放的同时有效降低CPU占用率,且该方法具有一定的实用性。     

一种基于多尺度实例学习的视频事件检测算法

现有的大多数视频事件检测方法首先从视频帧或视频快照中提取特征,然后对特征进行量化和汇集,进而为整个视频生成一个向量表示.最后的汇集步骤虽然简单高效,但是可能丢失时间局部信息,而这些信息对于确定长视频中事件发生的位置具有重要作用,从而削弱了事件检测的准确性.为此,本文首先将每个视频表示为多个“实例”,并将其定义为不同时间间隔的视频段.然后,针对每个视频的正实例比例已知和未知两种情况,提出基于多尺度实例学习的检测算法,在将实例标签看成隐藏潜在变量的同时推断出实例标签以及实例尺度的事件检测模型.最后,利用大规模视频事件数据集进行了全面的仿真实验,结果证明了本文算法具有显著的性能提升.此外,算法还可以确定视频中导致正检测的时间段的位置,进而对检测结果做出解释.     

一种新的快速局部不变特征算法

针对传统局部不变特征算子主方向提取不准确和匹配阶段过于耗时的问题,提出一种基于RI-LBP 算子和混合spill 树的快速局部不变特征算法。首先提出一种FAST-Difference 算法,提取出模板图像和待匹配图像的稳定特征点,然后使用旋转不变的RI-LBP 描述符计算特征向量,最后对特征向量集使用混合spill 树进行匹配并使用RANSAC 算法剔除误匹配点。RI-LBP 算子自身的旋转不变性能够在一定程度上克服特征点主方向确定不准确的缺点,使特征描述符的提取更加稳定,并生成更简单的53 维局部不变特征描述符。混合spill 树相对于kd-tree 省略了回溯过程,对于高维数据拥有更好的匹配效率。实验证明:该算法与SURF 算法描述能力相近,旋转和光照条件下比SURF 性能更优,并且匹配速度更快。     

基于快速尺度空间的无人机影像自适应稳像方法

为实现无人机航拍影像的实时稳像,针对稳像过程中特征检测的速度问题和运动滤波的发散现象,提出了一种改进的AGAST算法与自适应Kalman滤波相结合的实时稳像算法。对于无人机实时航拍视频序列,以当前帧的前一帧为基准进行稳像处理。改进的AGAST特征检测算法在尺度空间的基础上快速提取AGAST角点特征,用二进制描述符对其进行描述,然后用汉明距离匹配特征点。对于已获得的匹配特征点对集合,用RANSAC原则剔除误匹配点,再计算运动矢量。最后使用自适应Kalman滤波提取出运动矢量中的无人机主动扫描分量,进行运动补偿,获得稳像结果。实验数据使用标准测试图集和自己采集的无人机航拍视频,实验结果表明,所提算法在连续视频序列处理时效果显著、速度快,能够满足实时稳像的需求。     

融合热释电红外传感器与视频监控器的多目标跟踪算法

现有基于热释电红外传感器的多目标跟踪系统在目标之间距离较近或者轨迹相交的情况下存在着误差较大的缺点.针对此缺点,提出了一种新型的基于热释电红外传感器与视频监测器协同工作的多目标跟踪方案.该方案可以充分利用两种传感器的优势,弥补在目标跟踪中的不足.算法采用最小二乘法利用热释电信息进行定位,并通过从图像或热释电传感器信号的幅频特性中提取特征信息来校正联合概率数据关联算法的关联矩阵,有效避免了错误关联.实验表明,该方案在多目标交叉情况下跟踪误差仅为其它算法的八分之一到四分之一.     

2-D mesh-based video object segmentation and tracking with occlusion resolution

This paper integrates fully automatic video object segmentation and tracking including detection and assignment of uncovered regions in a 2-D mesh-based framework. Particular contributions of this work are (i) a novel video object segmentation method that is posed as a constrained maximum contrast path search problem along the edges of a 2-D triangular mesh, and (ii) a 2-D mesh-based uncovered region detection method along the object boundary as well as within the object. At the first frame, an optimal number of feature points are selected as nodes of a 2-D content-based mesh. These points are classified as moving (foreground) and stationary nodes based on multi-frame node motion analysis, yielding a coarse estimate of the foreground object boundary. Color differences across triangles near the coarse boundary are employed for a maximum contrast path search along the edges of the 2-D mesh to refine the boundary of the video object. Next, we propagate the refined boundary to the subsequent frame by using motion vectors of the node points to form the coarse boundary at the next frame. We detect occluded regions by using motion-compensated frame differences and range filtered edge maps. The boundaries of detected uncovered regions are then refined by using the search procedure. These regions are either appended to the foreground object or tracked as new objects. The segmentation procedure is re-initialized when unreliable motion vectors exceed a certain number. The proposed scheme is demonstrated on several video sequences.     

Hybrid Multicast and Segment‐Based Caching for VoD Services in LTE Networks

This paper proposes a novel video delivery scheme that reduces the bandwidth consumption cost from a video server to terminals in Long‐Term Evolution networks. This proposed scheme combines optimized hybrid multicast with a segment‐based caching strategy for use in environments where the maximum number of multicast channels is limited. The optimized hybrid multicast, allocation of multicast channels, and cache allocation are determined on the basis of a video's request rate, the related video's length, and the variable cost per unit size of a segment belonging to the related video. Performance evaluation results show that the proposed scheme reduces a video's delivery costs. This work is applicable to on‐demand TV services that feature asynchronous video content requests.     

Video object segmentation in rainy situations based on difference scheme with object structure and color analysis

A scheme based on a difference scheme using object structures and color analysis is proposed for video object segmentation in rainy situations. Since shadows and color reflections on the wet ground pose problems for conventional video object segmentation, the proposed method combines the background construction-based video object segmentation and the foreground extraction-based video object segmentation where pixels in both the foreground and background from a video sequence are separated using histogram-based change detection from which the background can be constructed and detection of the initial moving object masks based on a frame difference mask and a background subtraction mask can be further used to obtain coarse object regions. Shadow regions and color-reflection regions on the wet ground are removed from the initial moving object masks via a diamond window mask and color analysis of the moving object. Finally, the boundary of the moving object is refined using connected component labeling and morphological operations. Experimental results show that the proposed method performs well for video object segmentation in rainy situations.     

利用视频帧块的高效半盲鲁棒水印算法

针对视频水印算法鲁棒性差或需过多先验信息提高鲁棒性的问题,提出利用视频帧块、KAZE特征匹配和DCT变换方法实现半盲鲁棒视频水印算法。首先,利用用户注册信息作为水印并为之产生唯一的秘钥,基于该秘钥随机产生唯一的帧块并存入数据库;然后,将帧与帧块进行KAZE特征匹配并检测出嵌入区域,进而嵌入水印;最后,将帧块与扭曲视频进行KAZE匹配修正视频各帧并提取水印。仿真试验结果显示,在半盲的前提下,本算法具有高鲁棒性、水印提取过程同步性、版权纠纷解决和非法传播追踪的优点,且具有高应用价值。     

A Novel Blind Video Watermarking Scheme Based on Independent Dynamic Component

Video copyright protection has become important as it is possible to make unlimited copies of digital video without quality loss. Video watermark is the main method of video copyright protection. A novel blind video watermarking scheme is proposed in this paper, in which independent component analysis (ICA) is used to extract the common static component and relative dynamic component from the consecutive frames. And then a designed watermark is embedded into the wavelet coefficients of the obtained dynamic component according to the 4-neighboring-mean based algorithm. The binarization of the 3-level-approximate sub-image of the I-frame in each GOP (GOP-Group of Picture) is selected as the watermark for this GOP. Given such watermark, the scheme can easily resist copy attack and temporal synchronization. Besides, because the watermarks of different GOPs and their embedding positions in a GOP are different, the intra-video collusion attack can be resisted effectively and there is no possibility for inter-video collusion attack to succeed. The simulation shows the feasibility of this scheme, which can resist well temporal synchronization, copy attack, and collusion attack. Published online: February 2006