为了保证视频内容被合法用户接收,且接收到的视频内容没有受到帧丢弃、帧重组等恶意攻击,对接收到的视频内容安全信息进行远程取证。提出了基于可信平台模块TPM的视频安全远程取证,对该远程取证的结构和过程进行了分析。为防止中间人伪造远程取证响应,提出了基于Fourier-Mellin的视频内容哈希算法,采用TPM对视频帧的特征进行认证;同时为提高远程取证的效率,提出了用于视频安全远程取证下的可区分粒度的取证模式,分析了如何能够有效地在粒度下视频序列中寻找不可信视频帧。通过实验证明了该视频哈希算法的有效性,并验证了可区分粒度取证模式的特点。
In order to ensure the video can be received by the legal client without attack of frame dropping, frame rearranging, and so on, a remote attestation protocol of video security based on Trusted Platform Module (TPM) is proposed. The architecture of the remote attestation protocol is explained in detail. Video content hash algorithm which is based on Fourier-Mellin is put forward to authenticate the video frame contents in order to prevent the man-in-the-middle from fabricating the response of remote attestation. To improve the efficiency of remote attestation, an attestation mode of differentiable granularity is developed. The experiments prove the effectiveness of the video hash algorithm and also demonstrate the characteristics of the proposed attestation mode of differentiable granularity.