一种基于隐私保护的云端视频监控移动物体检测方法

A Moving Object Detection Method of Surveillance Video Based on Privacy Protection in the Cloud

随着云计算技术的快速发展,许多监控视频数据的存储和处理被外包到云端。然而,把原始数据直接发送到云端会威胁到我们的隐私安全。利用云计算技术在保证视频监控应用的同时隐私信息不被泄露,是目前亟待研究和解决的热点问题。针对这一问题,本文研究并实现了一种基于反色、置乱、矩阵乘的有效视频帧加密方法,使用者可以将加密后的视频保存在云端而不需要担心隐私泄露,同时在加密的视频中还可以有效地进行移动物体的快速检测。实验结果表明:该方法与在未加密的视频帧中进行移动物体检测具有相似的精度。

With the rapid development of cloud computing technology,more and more surveillance video data and business are stored and migrated to the cloud; however,the raw data directly sent to the cloud will be a threat to the privacy of users. How to make effective use of cloud computing technology and ensure that the video surveillance applications while private information will not be disclosed now becomes a hot issue. In order to solve the problem,an effective video encryption method is studied and implemented; user can save the video encrypted in the cloud without loss of privacy,while an encrypted video can also be used to detect moving objects. Experimental results show that the method has similar accuracy with detecting mobile object in unencrypted video frame.