Abstract:

The development of image hiding technology provides a new solution to the covert military communication.The
paper proposes a novel image hiding algorithm based on Wavelet Packets Transform（WPT） and Sigular Value Decomposition
（SVD）.The low frequency coefficient matrix of WPT has strong anti-jamming and the singular value is stability.Based on
the bit plane decomposition of the secret image，the bit planes with important information are hided into the singular value
matrix of the low frequency coefficient matrix，and cofficient matrixes of higher entrogy energy would be selected from the
remainder coefficient matrixes as the carries of the bit planes with lesser information.The hiding location is adaptively determined
by the carried image.The experiment results show that，not only the algorithm is robust against normal image processing
and attacks，such as filtering processing，adding noise，and cutting processing，but also robust against geometrical attacks，
such as rotation.It has good invisibility and big capability，too.

摘要： 图像隐藏技术的发展为军事隐蔽通信提供了新的实施途径。提出基于奇异值分解和小波包变换的图像隐藏算法，利用小波包变换低频子带抗干扰性强和图像矩阵奇异值的稳定性，对秘密图像做位平面分解并对载体图像做小波包分解，将包含重要信息的位平面隐藏在小波包分解低频系数矩阵的奇异值矩阵中，将包含次要信息的位平面隐藏在熵能量较大的其他剩余子带图像系数矩阵中，次要位平面的隐藏位置根据载体图像的内容自适应确定。实验表明，该算法可实现图像信息的安全隐藏，不仅具有较强的抗几何攻击的能力，且对高斯、剪切和滤波等攻击的鲁棒性也较好，同时具有较好的不可见性，隐藏容量大。