基于层间处理的可伸缩视频感知加密算法

Perceptual Encryption Algorithm of Scalable Video Based on Inter-layer Processing

为了提高可伸缩视频感知加密的安全性,提出了一种层间预测过程和层内纹理信息的联合加密算法。在编码层间预测残差信息前,加密层间颜色映射过程中所采用的映射系数,导致解码出错和颜色失真。在增强层的层内同时加密非零量化变换系数的绝对值和符号,进一步扩大加密空间,提高安全性。实验结果表明该算法保留了格式兼容特性,相比于现有的加密算法具有更高的视觉安全性,适用于不同可伸缩特性,计算复杂度低,比特率几乎保持恒定。

In order to improve the security of perceptual encryption for scalable video, a joint encryption algorithm for inter-layer prediction and intra-layer texture information is proposed. Before coding the inter-layer prediction residual information, it encrypts the mapping coefficients used in the inter-layer color mapping, leading to decoding errors and color distortion. At the same time, the absolute value and sign of non-zero quantized transform coefficients are encrypted in the enhancement layer, which further expands the encryption space and improves security. Experimental results show that the algorithm preserves format compatibility, has higher visual security than existing encryption algorithms, is suitable for different scalable features, has low computational complexity and almost constant bit rate.