Abstract: To provide larger capacity of the hidden secret data and to maintain a better visual quality of stego-image，a novel steganographic method based on SOM and HVS is presented.The contrast and texture sensitivity were used as feature vector，and by SOM，the pixels were classified into the sensitive class and the less sensitive class.More secret data could be embedded in the pixels of less sensitive class，while less secret data could be embedded in the pixels of sensitive class.From the experimental results，compared with SOC，the proposed method hides much more information and maintains a better visual quality of stego-image.

Key words: steganographic method, Self-Organizing Map（SOM）, Human Visual System（HVS）

摘要： 为了提供较大的秘密信息嵌入量和保持较好的载密图像质量，提出了一种基于自组织特征映射神经网络和人眼视觉特性的图像密写。该密写方法将对比度和纹理敏感度作为特征向量，并通过自组织特征映射神经网络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SOC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, SOM, HVS