基于灰色关联度的自适应图像多小波域密写

多小波变换解决密写问题的缺点是多比特嵌入容易破坏系数间的相关性,为了解决这个问题,首先运用灰色关联分析计算邻域内多小波树组的中心树序列与外围树序列的相关程度,根据相关性的大小自适应地控制系数的可替换比特数,并在LL、LH和HL频带系数进行比特替换。对多幅图像满嵌后的隐藏容量和含密图像PSNR值的实验结果表明,该算法作为变换域的隐藏方法容量较大,隐蔽性很好,实现了在满足隐蔽性的前提下进一步提高容量的目的。另外,抗攻击检验结果表明,从受到攻击的含密图像中提取的秘密图像的质量较好,具有一定的稳健性。

Steganography of self-adaptive images in multiwavelet domain based on grey correlation degree

The shortcoming of mul tiw avelet steganography is that mult i-bit s embedding can easily dest ro y the cor relat ion o f co ef ficient s. To solve the problem, grey relat io nal analy sis w as firstly used to calculate the cor relat ion deg ree betw een the cent ral t ree sequence and the peripheral t rees sequence of tr ees g roup in the adjoining region, the number of bit s w hich could be r eplaced w as self-adapt ively co nt rolled, and then the bit s replacement s to LL, LH and HL subband co ef f icient s were done. Ex periment s of capacity and PSNR for many fully embedded steg ano imag es indicate that the steganogr aphy has goo d impercept ibility and lar ge capacity as an information hiding method, and realizes the g oal of enhancing the capacity under the pr econdit io n of impercept ibility. In additio n, the anti-at tack test s show that the secr et images recover ed f rom the at tacked stegano imag es have high quality and the steganography has some solidity