摘要
基于Ni算法产生的图像失真的两个关键因素,提出一个新的改进算法.一方面,对图像块的类型进行预测,只允许在特定块内进行直方图平移,尽可能减少平移像素的总数,同时,对秘密信息进行预处理,尽量减少数据序列中1的比例,从而达到提高图像质量的目的.另一方面,采用双向直方图平移,一定程度上扩充了可隐藏数据的容量.实验结果表明,该算法不仅可以实现可逆数据隐藏,而且与Ni算法相比,隐藏数据后图像的失真更小.
An improved algorithm is proposed based on two key factors based on image quality of Ni＇s algorithm. At first, the type of image block is predicted. Histogram is only allowed to shift in particular image blocks, so the number of shifted pixels can be decreased. Secondly, the secret data will be pre-processed to get the less number of bit ＇ 1＇ , which can improve image quality. At the same time, the capacity of the hidden data is increased based on the two-direction shifting. The test results show that the algorithm can accomplish the reversible data hiding, furthermore, compared with Ni algorithm, it has lower distortion after information hiding.
出处
《信息与控制》
CSCD
北大核心
2013年第5期595-600,共6页
Information and Control
基金
国家自然科学基金资助项目(61070250)
中办信息安全重点实验室青年科研基金资助项目(YQNJ0904)
关键词
数据隐藏
可逆
直方图平移
低失真
data hiding
reversible
histogram shifting
lower distortion
