摘要：
根据常用隐写算法Outguess、F5、Steghide对JPEG图像量化DCT系数首位数字直方图和独立直方图的改变，提出一种JPEG图像隐秘检测算法。该算法提取JPEG图像量化DCT系数首位数字直方图和独立直方图作为图像特征，共144维，并采用Fisher分类器进行识别。实验结果证明，该算法检测率高、适用性好。
关键词:
信息隐藏,
隐秘检测,
JPEG图像
Abstract:
According to the changes of the historgrams of the first digital distribution and the individual DCT coefficients caused by the steganography algorithm, such as Outguess, F5, Steghide, this paper proposes a new detection algorithm for the JPEG Image. It extracts 144 features by using the first digital distribution and the histograms of individual DCT coefficients, and uses Fisher classifier for recognition. Experimental results show that the algorithm has high detection rate and chronic adaptability.
Key words:
information hiding,
steganography detection,
JPEG image
中图分类号:
张天祥, 潘峰, 李秀广, 王世峰. JPEG图像隐秘检测算法[J]. 计算机工程, 2010, 36(24): 125-126.
ZHANG Tian-Xiang, BO Feng, LI Xiu-An, WANG Shi-Feng. Steganography Detection Algorithm for JPEG Image[J]. Computer Engineering, 2010, 36(24): 125-126.