基于3D-DCT和混沌加密的医学体数据多水印算法

Multiple watermarks for medical volume data based on 3D-DCT and Logistic Map

含有病患信息的多媒体医疗数据在网络上存储、传输时,易受到恶意篡改或攻击,降低其安全性。基于三维DCT和混沌加密技术,提出了一种医学体数据鲁棒多水印算法。应用三维DCT变换提取医学体数据的视觉特征向量,结合Hash函数和第三方概念来进行多水印的嵌入和提取;使用混沌加密技术对多水印进行加密,提高其安全性。实验结果表明,该算法在提高水印容量的同时对常规攻击和几何攻击仍有很好的鲁棒性,实现了零水印嵌入和水印的盲提取,更好地保护了医学体数据。

When the multimedia medical data which contains patients＇ important information being stored and transmitted on the internet, it is vulnerable to be maliciously tampered and attacked. Thus, its security is quite low. A robust multiple watermarks algorithm for medical volume data is proposed based on 3D-DCT and Logistic Map. Applying 3D-DCT to obtain the visual featurevector of medical volume data, then, combining the Hash function and the third part concept, multiple watermarks are embedded and extracted accurately. The security of the multiple watermarks is improved after encryption by Logistic Map. The experimen tal results show that the proposed algorithm can embed more watermarks, and is robust against common and geometrical at tacks. Moreover, zero-watermarking embedding and blind extracting are realized in the proposed algorithm, which can protect the medical volume data better.