【作者】 尚宇雄；

【导师】 尚宇；

【作者基本信息】 西安工业大学 ， 通信与信息系统， 2011， 硕士

【摘要】 随着计算机网络和通信技术的发展,多媒体信息的应用越来越广泛,而由于数字信息容易被盗取和篡改,对数字信息的加密和保护成为了信息安全的重要课题。与此同时,分数阶Fourier变换是现阶段光学、信息学的研究热门,将分数阶Fourier变换应用到图像处理中也是一个重要的研究方向。论文重点研究了分数阶Fourier变换的基本理论和性质,和它在信号处理、图像分析和图像加密中的应用,并对现代图像加密技术中应用较为广泛的混沌加密技术做了研究,分别分析了混沌置乱加密技术和变换域加密技术的特点,并针对现有加密技术的不足,提出了结合混沌置乱和分数阶Fourier变换的多重图像加密算法。本文对基于分数阶Fourier变换的单、多分量chirp信号检测、参数估计及分离,含噪LMF信号滤波进行了分析讨论,仿真结果表明分数阶Fourier变换相较于Wigner-Ville分布和短时Fourier变换而言能够更好的分离和检测多分量chirp信号,对于含有加性噪声的单分量chirp信号,在信噪比较低时依然有较好的滤波效果；根据分数阶Fourier变换系数对阶数的敏感性和多样性,提出了基于分数阶Fourier变换的图像加密算法并仿真,结果表明该算法拥有较好的鲁棒性,能更好的抵抗统计特性攻击；在对混沌置乱图像加密技术做了一些研究后,提出基于混沌置乱的加密算法并仿真,该算法拥有较大的密钥空间,安全性好。通过对混沌置乱图像加密算法和分数阶Fourier图像加密算法进行仿真实验和分析对比,确定了优化加密算法的目的和思路,提出了改进的图像加密算法。仿真结果显示密文与明文相比灰度直方图更加平坦,均方误差与密钥极为敏感,相关性极低；算法具有较强的抗遮盖剪裁,抗噪声性能。最后,在DSP6711的平台上实现了加密算法,加密图像完全掩盖了明文的空频域信息。更多还原

【Abstract】 As the rapid development of computer networks and communication technology, multimedia information is more widely used. Because digital information is easy to be stolen and tampered, the encryption and protection of digital information has become an important issue in information security field. Meanwhile, fractional Fourier transform is hot research topic both in optics and informatics, it is also a research direction in image processing. This paper studied the nature of fractional Fourier transform and its applications in signal processing, image analysis and image encryption. The application in modern image encryption of chaotic technology was studied. An image encryption algorithm was proposed after the analyzing the characteristics of the chaotic scrambling encryption technology and transform domain encryption technology.This paper also discussed the detection, parameter estimation and separation about the single and multi-component chirp signal, and the filtering of LMF signal with noise based on fractional Fourier transform. Simulation results demonstrated that fractional Fourier transform can separate the multi-component chirp signal well compared with Wigner-Ville or Short-time Fourier transform, also, the single component chrip signal with additive noise still has a better filtering effect even though the SNR was low. This paper proposed and simulated an image encryption algorithm based on fractional Fourier transform by using its coefficient’s sensitivity and diversity to initial value. This algorithm has good robustness and better ability to against statistical properties’attacks. An encryption algorithm based on chaotic scrambling was proposed and simulated, the algorithm has larger key space and better safty, but it seldom prevent clipping or noise, its robustness is poor in addition. An improved image encryption algorithm was present after analyzing the results of chaotic scrambling image encryption algorithm and fractional Fourier image encryption algorithm simulation experiment in this paper. The result show that the histogram of cryptograph became more flatter, the MSE and relativity was sensitived to the key. At last, the paper realized encryption algorithm on DSK6711. The result shows that the space and frequency information of original image were hidden perfectly.更多还原