Image Parallel Encryption Algorithm Design And The Application For Mobile Terminal Equipment

In the era of big data,digital information such as texts,voices,images,and videos are widely transmitted on the Internet and various mobile terminal devices,which has revolutionized people's lives,work,and study.While the development of the Internet and digital multimedia technology has provided many conveniences,the data security problem has also come along.As an important carrier of digital information,digital images are inseparable from people's daily life.Compared with other digital information,digital images have the characteristics of large amount of data,high redundancy,and high correlation between adjacent pixels,which leads to the traditional encryption methods such as AES,DES,and IDEA.Its information protection is not ideal.Because of its excellent pseudo-randomness,ergodicity,initial sensitivity and unpredictability,chaos systems have attracted extensive attention of researchers.Nowadays,chaotic systems have gradually been applied to digital information encryption.Of course,the protection of digital images by chaos also shows its outstanding advantages,making the protection of image information through chaotic systems an important research hotspot.In recent years,researchers have proposed many chaotic digital image encryption schemes,but due to the flaws in the design of some encryption structures,they are vulnerable to attack by some deciphering algorithms.In addition,due to the continuous improvement of the accuracy of image sensors,the amount of image data is also increasing.In the face of massive digital images,people need to find an efficient and secure image encryption solution.For some common chaotic image encryption schemes are often calculated by the CPU serial.Because the CPU has a small number of calculation units,if the image pixels are sequentially executed by the CPU and a simple encryption algorithm is executed,the calculation time required is longer.Compared to CPUs,GPUs have a greater number of computational units,although they have less computing power than CPUs.Therefore,the parallel computing power of the GPU is much stronger than that of the CPU.With the continuous improvement of hardware technology,GPU parallel computing capabilities have also been highlighted.If you can design a parallel image encryption structure and be able to accelerate it in parallel through the GPU,its encryption performance will be effectively improved.Therefore,by parallelizing the image encryption algorithm on the GPU platform to accelerate image encryption,the real-time transmission of image information can be guaranteed.Based on the existing research results,based on the consideration of the structure of the image encryption system,the algorithm is implemented in parallel.The OpenCL and Metal parallel programming languages and frameworks are used to implement the encryption algorithms on the GPUs of the PC and handheld devices respectively.Parallelization.The main work of this article is reflected in the following four aspects:1.First,the research background of the topic is introduced,including chaotic image encryption and GPU parallel computing theory,and then several innovations of this paper are given.We described the basics of chaos theory and the structure of digital image encryption system structure.The current research status of traditional chaotic image encryption,parallel image encryption and GPU parallel image encryption are briefly summarized.Finally,it analyzes several deficiencies in digital image encryption and gives corresponding solutions.2.Aiming at the problem of ignoring the correlation between image layers in traditional image encryption schemes,an adaptive color image interactive encryption scheme is proposed.After different pixel layers of a color image are quantized by a quantization scheme,an initial quantification decimal fraction with plaintext characteristics is obtained.As tent tentative initial values and initial parameters,and to obtain the corresponding chaotic sequence,the image is scrambled by the new Rubik's cube scrambling algorithm.After that,the relevant eigenvalues of the scrambled image are obtained as the relevant parameters of the piecewise linear chaotic system to generate the corresponding diffusion sequences,and the encrypted image is obtained through the interleaving diffusion method.Theoretical analysis and experimental results show that the algorithm can improve the coding efficiency,enhance the security of ciphertext,have a larger key space and higher key sensitivity,and can resist statistical and exhaustive attacks.3.A parallel image encryption scheme was designed using piecewise linear chaos(PWLCM)and four-dimensional hyperchaotic system(FDHCM).First,using the new parallel quantization method,two decimal numbers based on the plaintext image and the external key are obtained.They are used as initial values and control parameters for PWLCM.Then,PWLCM and FDHCM,which use them to control the scrambling and diffusion processes,construct an encryption matrix and four chaotic sequences.The proposed algorithm is implemented and tested in parallel on a graphics processing unit(GPU)device device.Numerical analysis and experimental results show that the algorithm achieves high encryption speed and good security performance,providing a potential solution for real-time image encryption applications.4.Implement a new type of image parallel encryption algorithm on handheld devices.the initial information based on the plaintext image,and it as an input for the parallel quantization scheme,the quantized decimals to be generated.Combined with the chaotic system,the chaotic scrambling and diffusion matrix are generated,and the image is encrypted by the “confusion and diffusion” structure,and the algorithm is implemented on the IOS handheld device through the Metal parallel framework.Experimental results show that the system can effectively improve the efficiency of image encryption,has a high degree of sensitivity to the plaintext,is very suitable for secure real-time communication of image data,but also provides a better solution for parallel processing of image data through the GPU.