IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 9, NO. 2, FEBRUARY 2014 309 Perceptual Encryption of H.264 Videos: Embedding Sign-Flips Into the Integer-Based Transforms Bing Zeng, Senior Member, IEEE , Siu-Kei Au Yeung, Member, IEEE, Shuyuan Zhu, Member, IEEE, and Moncef Gabbouj, Fellow, IEEE Abstract— An alternative-transforms-based scheme has recently been proposed to achieve perceptual encryption of video signals in which multiple transforms are designed by using different rotation angles at the final stage of the discrete cosine transforms (DCTs) butterfly flow-graph structure. More recently, it is found that a set of more efficient alternative transforms can be derived by introducing sign-flips at the same stage, which is equivalent to an extra rotation angle of π . In this paper, we generalize this sign-flipping technique by randomly embedding sign-flips into all stages of the DCTs butterfly structure so that the encryption space becomes much larger to yield a higher security. We pursue this study for H.264-compatible videos, assuming that the integer DCT of size 4 × 4 is used. First, we follow the separable implementation of the 4 × 4 2-D DCT in which different sign-flipping strategies will be employed along its horizontal and vertical dimensions. Second, we convert the 4 × 4 2-D DCT into a 16-point 1-D butterfly structure so that more sign-flips can be embedded at its various stages. Third, we choose different schemes to pair the node-variables in the 16-point 1-D butterfly structure, thus further enlarging the encryption space. Extensive experiments are conducted to show the performance of these improved encryption schemes and some security analyzes are also presented to confirm their persistence to various attacking strategies. Index Terms— Video encryption, perceptual video encryption, integer-based transforms, H.264/AVC. I. I NTRODUCTION T HANKS to the modern networking technologies, par- ticularly the high-speed Internet, we can now exchange information with each other in any place and at any time. However, the security issue rises to become an important one we need to consider such that no illegal parties can obtain useful information. In this work, we focus on the encryption-based solutions to the security problem for video Manuscript received June 21, 2013; revised September 16, 2013; accepted November 23, 2013. Date of publication December 3, 2013; date of current version January 14, 2014. This work was supported by the National Natural Science Foundation of China under Grant 61370148. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. H. Vicky Zhao. B. Zeng and S. Zhu are with the Institute of Image Processing, School of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China (e-mail: eezeng@ust.hk; ecezhshy@gmail.com). S.-K. Au Yeung is with the School of Professional and Continuing Educa- tion, The University of Hong Kong, Hong Kong (e-mail: eejeffay@ust.hk). M. Gabbouj is with the Department of Signal Processing, Tampere Univer- sity of Technology, Tampere 33720, Finland (e-mail: moncef.gabbouj@tut.fi). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIFS.2013.2293955 signals-the fastest growing information that is being exchanged over today’s networks. Being regarded as one type of data, video signals can also be encrypted by some conventional cryptographic techniques such as the data encryption standard (DES) or advanced encryption standard (AES) [1] that have been widely used for text data. The problem however is that these conventional techniques usually require a long processing time, thus not suitable for video signals that typically have a huge volume and need to be processed in real-time. Unlike many types of data, a visual data (including images and videos) would be deemed as useless if its visual contents are lost. Based on this feature, a number of algorithms with a much reduced compu- tational cost have been developed to encrypt only a selected set of significant information before, during, or after the video compression. Clearly, encryptions that are carried out before or after the video compression are compression-independent. Examples of this category include the correlation-preserving permutations [2], the puzzling algorithm [3], and some algo- rithms that are applied in the compressed bit-stream domain [4], [5]. On the other hand, encryption can also be embedded into a particular stage within the video encoder, leading to the compression-dependant algorithms. Examples include the zigzag scanning permutation [6], the real-time video encryp- tion algorithm (RVEA) [7], [8], the use of multiple Huffman tables (MHT) [9]–[11], and a joint signal processing with encryption approach [12]. Based on applications, a video sequence may need to be encrypted completely so that zero visual information can be reconstructed by an un-authorized person, no matter how hard this person tried. These applications include video conferenc- ing and video telephony, as well as video transmission for financial or military purposes. We name them as the “fully confidential” video encryption algorithms. Many previously developed algorithms can achieve this goal, including the MHT-based ones, various encryption schemes over com- pressed bit-streams, and encryptions on intra-prediction modes and/or motion vectors. On the other hand, we believe that perhaps a more popular type of applications is video transmis- sion for entrainment, such as video-on-demand (VoD), pay-TV, and live video broadcasting. In this scenario, the encryption criteria are two-fold: i) the video signal is encrypted to a certain extent so that only customers who have paid for the service can obtain the high-quality version; and ii) some video contents are still visible, but at a much lower quality, for 1556-6013 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.