As video editing software becomes more and more powerful,video content tampering often occurs,and the destruction of video data (including frame deletion,cutting,editing) causes the video information unreliable.In order to detect high-definition video tampering,the video-related information (including user name,video device number,timestamp,frame sequence number,hash value) is added into video stream during video encoding.By using the temporal and spatial correlation characteristics of video adjacent pixels,the appropriate embedding algorithm is adopted to realize the information hiding in the video compression field.In the video decoding stage,the hidden information is extracted with mapping rule corresponding to the embedding scheme,and the hash value of the extracted information is recalculated and compared with the original hash value to judge the integrity of the video.The experimental results show that the proposed scheme keeps HD video quality well and can effectively detect video integrity threats such as frame editing,deletion and replacement,thus has good application significance.%随着视频编辑软件的日益强大,视频内容篡改问题时常发生,对视频数据的破坏(包括帧删除,剪切,编辑)造成了视频无法提供真实可靠的参考信息.课题针对视频内容完整性问题进行方案研究,利用视频邻近像素的时空相关特性,采用合适的嵌入算法,将视频相关信息(包含用户名,视频设备编号,时间戳,帧序列号,哈希值)隐藏在视频编码压缩域中.在视频解码阶段,按照嵌入方案对应的映射规则将隐藏信息提取出来,重新计算提取信息的哈希值,并与原始的哈希值比对,根据比对结果判断视频信息完整性是否遭到破坏.实验结果表明该方案不仅对视频质量没有观感上的影响,同时能够有效地检测帧编辑、删除、替换等视频完整性威胁,实现了高清视频内容完整性检测,具有良好的应用意义.
展开▼
