基于阈值划分的加密域信息隐藏算法  

Reversible data hiding in encrypted domain based on threshold

作　　者：张巧妹[1,2] 顾军华[3,2] 侯向丹[3,2] 

机构地区：[1]河北工业大学电子信息工程学院,天津300401 [2]河北工业大学河北省大数据计算重点实验室,天津300401 [3]河北工业大学计算机科学与软件学院,天津300401

出　　处：《计算机工程与设计》2017年第6期1424-1428,1433,共6页Computer Engineering and Design

基　　金：天津市自然科学基金项目(16JCYBJC15600)

摘　　要：为解决图像中隐秘信息的嵌入量和提取秘密信息的准确性问题,增强信息的安全性和可靠性,提出一种基于阈值分类图像块相关性的加密域信息隐藏算法。利用分块图像中相邻块的边界像素的相关性特征,通过选取合适的阈值对图像进行纹理区域和平滑区域划分;根据嵌入内容和区域划分结果,将加密之后的图像像素值以两种低位(不重要位)取反的方式进行隐藏;接收者能较好地提取隐藏的数据,恢复原始图像。大量实验结果表明,该算法能有效降低数据提取和图像恢复的错误率,嵌入容量比之前算法提高了一倍左右。According to the safety and reliability of the information,to improve the amount of confidential information embedded in the image and the accuracy of confidential information extraction,an improved information hiding algorithm in encrypted domain based on the classification of correlation of image blocks was proposed.The image block was divided into smooth region and texture region by selecting appropriate threshold.The result of the division of correlation characteristics of adjacent boundary pixel block was used to hide the information in the way of flipping the two numbers of LSBs(least significant bits).The ultimate recipient can better extract the hidden data and recover the original image.A large number of experimental results show that the improved algorithm can effectively reduce the error rate of data extraction and image restoration,and its embedding capacity is about twice larger than that of the original algorithm.

关 键 词：加密域 可逆信息隐藏 相关性 阈值 嵌入容量 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]