目的实现数字图像与QR二维码信息的安全保密。方法基于频域-空域双重加密的思想,提出Haar整数频域变换耦合动态引力模型的图像无损加密认证算法,并将该算法应用于QR二维码的加密传输。首先引入Haar小波变换,定义频域系数修整模型,将明文分解为4个子带;随后基于256位外部密钥,迭代3D Chen系统,建立混沌序列择取与优化机制,有效消除瞬态效应,从而输出3个优化子序列,通过融合这些序列,输出一组密钥流,基于升序排列,形成位置扰乱源,对4个子带完成频域置乱,再利用Haar逆变换,形成置乱密文;构建像素点质量动态估计模型,改进引力模型,对置乱密文完成空域扩散;定义密文深度分段扩散机制,对初始密文完成二次扩散,提高密文的NPCR(Number of Pixels Change Rate)与UACI(Unified Average Change Insensitive)值;最后,引入HASH检测机制,赋予算法决策功能,对图像在传输中是否遭到攻击进行认证。结果与基于混沌理论的加密技术相比,文中算法具备更高的安全性与抗剪切攻击能力,且解密图像的失真度最小;同时,对QR二维码信息也具有较高的保密度和较低的解密失真度,在安全加密的同时,也较好地保留了QR二维码的原有结构信息。结论文中算法具有较高的安全性,能够安全保护图像与QR二维码在网络中安全传输,在包装与印刷防伪条码领域具有较好的实际应用价值。
The work aims to realize the security and confidentiality of the digital image and the QR code information. The verified image non-destructive encryption algorithm of transforming coupling dynamic gravity model based on Haar integer frequency domain was proposed based on the thought of double encryption of frequency domain - spatial domain. Such algorithm was used for encryption and transmission of QR code. Firstly, the plain texts were decomposed into four sub-bands by introducing wavelet transformation and defining dressing model of frequency domain coefficient. Then 3D Chen system was iterated and chaotic sequence selection and optimization mechanism were built based on the 256-bit ex- ternal key to effectively eliminate transient effects, thus outputting 3 optimized sub-sequences. Through the fusion of these sequences, one group of key streams were outputted and the position disturbance source was formed according to the ascending order to enable the four sub-bands to be scrambled in frequency domain. Then the permutation cipher was ob- tained by using the Haar inverse transformation. The permutation cipher was diffused based on the improved gravity model by construction of dynamic pixel quality estimation model. The original cipher was diffused again by defining the cipher depth diffusion mechanism for upping the NPCR (Number of Pixels Change Rate) and UACI (Unified Average Change Insensitive) value. Finally, whether the image was attacked in transmission was identified by introducing the HASH detection mechanism to endow the algorithm with function of decision-making. The algorithm herein had a higher security and anti-shear ability, and the distortion of the decrypted image was the least compared with the encryption technology based on chaos theory. At the same time, the QR code informatioia also has a higher confidentiality and a lower decryption distortion, and to a better extent retains the original structure of the QR code information. The algorithm herein has a higher security and can safely p