摘要
在一些保密要求严格的场合,监控视频泄露个人隐私,已引起了人们的广泛关注。针对视频中车牌区域加密的问题提出一种基于H.264的车牌快速保护方案。首先使用HSV颜色模型搜索出车牌区域,并通过H.264中的灵活宏块排序FMO(Flexible Macroblock Orde)技术将一帧中车牌区域和背景区域的宏块映射到不同的组。然后提出一个一维混合混沌序列,用来生成二值密钥流。最后在参考适应性二元算术编码CABAC(Context-based Adaptive Binary Arithmetic Coding)编码过程中使用该密钥流对车牌区域进行加密。实验结果表明,该方案不仅具有良好的鲁棒性,而且快速有效地保护了车牌区域。
The monitoring videos often leak the privacy in some occasions where the secrecies are strictly demanded,which has attracted wide attention.Aiming at the encryption of licence plate area in videos we propose an H.264-based rapid license plate protection scheme. First,we use HSV colour model to search the area of license plate,and employ FMO （flexible macroblock order）technique in H.264 to map the macroblocks of license plate area and background area in one video frame into different groups.Then,we present a one-dimensional mixed chaotic sequence to generate a binary key stream.Finally,we use the key stream to encrypt license plate area in the process of context-based adaptive binary arithmetic coding （CABAC）.Experimental results show that the scheme has good robustness and also protect the license plate area rapidly and effectively as well.
出处
《计算机应用与软件》
CSCD
2015年第10期306-310,333,共6页
Computer Applications and Software
基金
国家自然科学基金项目(61001004)
湖南教育厅基金项目(11B002)
湖南省海外名师项目(2013008)
湖南省研究生科研创新项目(CX2013B376)
