摘要
本文将图像进行分块,根据每块图像量化后的DCT系数矩阵,结合视觉特性将图像块分为不同的类型;然后对不同类型的块采用不同的方法在AC系数中嵌入秘密信息,减少对载体图像的影响,隐藏信息后载体图像块的类型不变,从而可正确提取隐藏信息。实验结果表明,该算法在保持JPEG图像视觉质量的同时有较高信息隐藏容量。
The cover image is firstly divided into non-overlapping blocks,which are then classified into different categories according to quantized DCT coefficients of each block and the characteristics of Human Visual System.Different methods are adopted to hide the secret data in AC coefficients for different types of blocks in order to reduce the influence on the cover image.The types of image block remain unchanged after secret is hidden in these blocks,thereby the secret data can be extracted exactly.Experimental results show that the proposed algorithm can not only maintain visual quality of JPEG image but also achieve high capacity.
出处
《广西师范大学学报（自然科学版）》
CAS
北大核心
2012年第3期119-124,共6页
Journal of Guangxi Normal University:Natural Science Edition
基金
国家自然科学基金资助项目(60963008)
广西自然科学基金资助项目(2011GXNSFD018026
0832104
2012GXNSFBA053166)
广西科学研究与技术开发计划项目(10123005-8)
重庆市教委科技项目(KJ121310)
涪陵区科技计划项目(FLKJ
2012ABA1056)
广西教育厅项目(200911MS55)
