J. Shanghai Jiaotong Univ. (Sci.), 2014, 19(4): 385-391 DOI: 10.1007/s12204-014-1513-7 Region of Interest Based Selective Encryption Scheme for Privacy Protection in H.264 Video GUO Jie ∗ (郭 捷), XU Jia-yun (徐家运), BAO Jia-li (鲍佳立) (College of Information Security, Shanghai Jiaotong University, Shanghai 200240, China) © Shanghai Jiaotong University and Springer-Verlag Berlin Heidelberg 2014 Abstract: A selective encryption scheme for region of interest (ROI) of H.264 video is proposed to protect the personal privacy in a video. The most important part of video can be protected with less cost and operation by only encrypting the content of ROIs. Human face regions are selected as ROI and detected by using Gaussian skin color model. Independent ROI encoding is realized with the mechanism of flexible macro-block ordering (FMO). Frames are divided into grid-like slice-groups which can be combined flexibly to form a required ROI. Both luminance component and chrominance component of the macro-blocks in ROI are modified to achieve good encryption quality and location accuracy. In the process of decryption, the encrypted area is located automatically. There is no need to transmit additional position information of ROIs to the end of decryption. The encrypted video is decrypted correctly with the secret key. JM18.4 software is employed to perform the simulation experiment. Experimental results show the accuracy and effectiveness of our scheme to encrypt and decrypt the ROIs in H.264 video. Key words: region of interest (ROI), selective video encryption, flexible macro-block ordering (FMO) CLC number: TP 309.2 Document code: A 0 Introduction H.264 is one of the most popular video coding stan- dards now, and it is able to keep a good video qual- ity at a low bit rate. With the diverse application of H.264 technology, video encryption methods are needed for the privacy protection of the video content. Some elements should be considered for designing an effec- tive video encryption scheme, such as security, com- pression, compatibility, complexity and preserved func- tionality. In the literature, many video encryption al- gorithms have been proposed [1] . However, most algo- rithms encrypt the whole frame, which will increase the computational cost. The region of interest (ROI) is considered as a valu- able and important area in a video, which requires to pay more attention. Only important part of the video content is encrypted so that low-complexity and high-security performance can be achieved. Newton et al. [2] proposed a privacy preservation scheme for de- identifying faces in an image. Qi et al. [3] proposed a face protection scheme for H.264 video. The face region is located with face detection technology and encrypted Received date: 2013-10-30 Foundation item: the National Natural Science Founda- tion of China (No. 61073157) *E-mail: guojie@sjtu.edu.cn based on context-based adaptive binary arithmetic cod- ing (CABAC). Kim et al. [4] proposed a selective video encryption for the ROI in scalable video coding. The ROI is limited to be a fixed area in different frames by setting the parameter in the picture parameter set (PPS) unit. Dufaux and Ebrahimi [5] proposed a scram- bling scheme for privacy protection in video surveil- lance systems. Each sequence needs a ground truth seg- mentation mask defining ROI. Zhao et al. [6] proposed an object-based unequal encryption method for H.264 compressed surveillance videos. Compared with equal encryption method, computational cost is still too high for real time applications, although the encrypted data ratio is reduced. Peng et al. [7] proposed an ROI pri- vacy protection scheme for H.264 video based on flexi- ble macro-block ordering (FMO) and chaos. FMO type 6 is used to define the ROI in the video. But there is no attention paid on the location of encrypted area in de- cryption. Because ROIs are variable in different frames, it is necessary to locate the ROI area correctly in de- cryption. Transferring the position information of ROIs to the decoder will increase the overload. Another way to locate the ROI position is to use information hiding algorithm. But the detection on watermark is still not accurate enough under attack. In this paper, a new selective encryption scheme is proposed to protect the face area in a video. Face area