无定位图的预测误差差值扩展可逆数据隐藏  

Reversible data hiding using prediction error difference expansion without location map

作　　者：熊志勇[1] 蒋天发[1] 

机构地区：[1]中南民族大学计算机科学学院,武汉430074

出　　处：《计算机应用研究》2010年第3期1015-1018,1028,共5页Application Research of Computers

基　　金：国家民委重点科研基金资助项目(MZY02004)

摘　　要：将Tian差值扩展技术应用于彩色图像中,提出一种利用预测误差差值进行扩展嵌入的彩色图像可逆数据隐藏算法。针对传统差值扩展技术存在过分修改像素灰度值、须嵌入定位图等缺点,首先利用色彩分量间的相关性减小差值,并将差值扩展量分散到两个色彩分量中;其次,对直方图平移技术进行改进,使得同等嵌入率下图像质量达到最佳;最后由两个色彩分量中像素的预测值之和决定可用于扩展嵌入的像素,无须保存溢出定位图,提取端在提取信息时可无损地恢复原始图像。实验结果表明,与其他算法相比,该算法在同等嵌入率下可取得更好的图像质量,算法复杂度更低。This paper proposed a reversible data hiding algorithm for color images using prediction error difference expansion. To avoid the drawback of Tian＇ s algorithm which must embed a location map and the quality decline of stego-images from excessive modulation to pixels, this algorithm using correlation of color components to decrease difference and disperse smaller expansion to two components, and improved histogram shifting technique to improve the quality of stego images, the embedding pixels was determined with sum of prediction values, so the algorithm did not need location map. Experimental results show the quality of stego-image is significantly improved, and the complexity is lower, when compared with other new or classical algorithms.

关 键 词：可逆数据隐藏 预测误差差值扩展 直方图平移 溢出定位图 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]