论文部分内容阅读
【目的】设计大数据环境下安全性和实时性更好的半脆弱水印算法,保护图博档视频资源的真实性和完整性。【应用背景】保证视频资源对常规视频操作的鲁棒性,满足图博档视频资源内容认证的实时性需求。【方法】运用量化调制的方法嵌入二值水印图像进行版权验证,在视频关键帧中嵌入索引水印检测帧间篡改,对视频帧的最低有效位进行异或运算生成认证水印检测帧内篡改。【结果】水印算法可以对视频资源进行有效的版权验证和内容认证,透明性较好,对于常规视频操作具有较强的鲁棒性,PSNR保持在33以上。篡改定位时间5s左右,具有良好的实时性。【结论】本研究有助于图博档视频资源的真实性和完整性保护,促进大数据环境下图博档信息资源共享和服务融合。
[Objective] To design a semi-fragile watermarking algorithm with better security and real-time performance in big data environment, and to protect the authenticity and integrity of the video resources of the map file. [Application Background] To ensure the video resource is robust to conventional video operations and meet the real-time requirements of content authentication of the video content in the picture file. 【Method】 The method of quantization and modulation was used to embed binary watermarking image for copyright verification. In the video keyframe, the index watermark was embedded to detect inter-frame tampering, and the least significant bit of the video frame was XORed to generate the certified watermark to detect the tampering in the frame. 【Result】 The watermarking algorithm can effectively validate the video resources and authenticate the content. The transparency is better and the robustness is more robust to the conventional video operations. The PSNR remains above 33. Tampering positioning time 5s or so, with good real-time. 【Conclusion】 This study helps to protect the authenticity and completeness of the video resources of Tuobo file and promote the sharing of information resources and service integration of Tuobo files in big data environment.