HEVC帧内预测模式和分组码的视频信息隐藏  

Video information hiding in intra prediction and block codes for HEVC

作　　者：王家骥[1] 王让定[1] 李伟[1] 徐达文[1] 徐健[1] 

机构地区：[1]宁波大学纵横智能软件研究所,浙江宁波315211

出　　处：《光电子．激光》2015年第5期942-950,共9页Journal of Optoelectronics·Laser

基　　金：国家自然科学基金(61170137;61301247;61300055);浙江省自然科学基金(LY13F020013);宁波市自然科学基金(2013A610059);浙江省"信息与通信工程"重中之重学科(XKX11405)资助项目

摘　　要：为了在最新的视频压缩标准-高效视频编码(HEVC)下实现信息隐藏并在保证隐秘信息嵌入容量的前提下减少对宿主信息的修改,提出基于帧内预测模式和分组码的HEVC视频信息隐藏方法。首先,利用分组码的标准阵列译码方法,建立(4,3)码标准阵列译表;然后,根据译码表与预测模式的映射关系,调制帧内4×4亮度块的预测模式嵌入隐秘信息;最后,对调制后的4×4亮度块重新编码,使得在连续4个4×4帧内亮度块嵌入3bit隐秘信息,平均修改1.25位预测模式,减小因调制预测模式对视频造成的影响。实验结果表明:所提算法的PSNR值下降在0.05dB以内,码率增长在1%左右,算法能很好的保证视频主客观质量,对视频的编码比特率影响很小。At the premise of ensuring embedding capacity, to reduce the modification of host for high effi- ciency video coding （HEVC） information hiding,which is the latest video compression standard , an in- formation hiding algorithm for HEVC based on intra prediction and block codes is proposed. In the first place, we establish a （4,3） standard array decoding table with the principle of the standard array. Then we modify the intra prediction modes in 4 X 4 luminance blocks according to the mapping between the matrix and the prediction mode. Finally, re-encode the changed block with the substitutive prediction mode. It makes that four 4 X 4 luminance blocks embed 3 bit information and modify 1.25 prediction modes on average, which reduces the impact of the prediction mode modulation on the quality of the vid- eo. Experimental results show that the bit-rate increases by about 1 % and the peak signal-to-noise ratio （PSNR） loss is just about 0. 05 dB. The proposed hiding algorithm can effectively embed information with little influence on the bit-rate and quality of the cover video.

关 键 词：信息隐藏 高效视频编码(HEVC) 帧内预测 分组码 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]