摘要： 为了提供较大的秘密信息嵌入量和保持良好的载密图像质量，该文提出一种基于模糊C均值聚类和人眼视觉特性的图像密写。该密写方法将对比度和梯度敏感度作为特征向量，并通过模糊C均值聚类将像素分为视觉敏感类与视觉不敏感类，将较多秘密信息嵌入属于视觉不敏感类的像素，而将较少秘密信息嵌入属于视觉敏感类的像素。实验结果表明，与四边匹配算法相比，该算法有更大的嵌入量并保持了良好的载密图像质量。

关键词: 密写, 模糊C均值聚类, 人眼视觉特性

Abstract: In order to provide larger capacity of the hidden secret data and to maintain a good visual quality of stego-image, a novel steganographic method based on FCM and HVS is presented. The contrast and gradient sensitivity are used as feature vector, and by the way of FCM, the pixels are classified into the sensitive class and the less sensitive class. More secret data can be embedded in the pixels of less sensitive class, and less secret data can be embedded in the pixels of sensitive class. Experimental results show , compared with the four-sided side match method, the method hides much more information and maintains a good visual quality of stego-image as well.

Key words: steganographic, Fuzzy C-means Clustering(FCM), Hunan Visual System(HVS)

中图分类号: 

• TP309.2