The Research Of CUDA-based Parallel Digital Image Encryption Algorithm

The digital image has the characteristics of large data redundancy and rich information,and the traditional encryption algorithm is not designed for these features.It can not complete the image encryption task safely and efficiently.Therefore,it is necessary to design the encryption algorithm with image characteristics.On the other hand,with the rise and development of GPU general purpose computing technology,the use of GPU resources for image parallel processing has become a hot spot.Based on this background,this paper studies the digital image encryption technology,and proposes two parallel image encryption algorithms.Specific research is as follows:(1)An image encryption algorithm based on spreading code is proposed.The algorithm applies the idea of constructing a block cipher for the diffusion code in image encryption.Firstly,the concept of diffusion code infrastructure is introduced.It uses the combination of diffusion code and Boolean function to realize cell diffusion and mixing process.It has the advantages of fast diffusion speed and few iterations.And then use the diffusion code infrastructure as the F function in the Feistel cipher structure to complete the image encryption process through three rounds of iteration.The experimental results show that the algorithm has the advantages of fast encryption speed and large key space,and can effectively resist the attacks such as statistical analysis attack and exhaustive attack.(2)An image encryption algorithm based on double chaotic system is proposed.The algorithm designs a parallel chaotic key generation algorithm,parallel scrambling algorithm and parallel diffusion algorithm and is used in image encryption.Firstly,after analyzing the problem of small key space and low security in single chaotic system,the Logistic map generation control parameters are proposed as the initial values of Skew Tent mapping,and the iterative generation of key sequences is carried out.The experimental results show that the double chaotic system is good Randomness.And then use the generated key sequence,the use of divide the strategy of the image line(column)scrambling and row(column)diffusion,complete the entire image encryption.Experiments show that the key sequence generated by the double chaotic map can effectively improve the key sensitivity of the algorithm,and the algorithm has strong resistance to exhaustive attack,statistical analysis anddifferential attack.(3)The above encryption algorithm in the CUDA platform parallel optimization.Based on the parallel characteristics of algorithms and data,the factors such as data organization and storage scheme are optimized on the basis of experiment.The efficiency of the optimized algorithm is obviously increased.The first algorithm is nearly 180 times faster than that of CPU platform.The second algorithms also have40 times faster than the speed.In this paper,considering the characteristics of image data,the characteristics of CUDA platform and the speed and security requirements in encryption algorithm,two secure and efficient image encryption algorithms are designed,which can perform real-time transmission of image or video information in parallel devices.At the same time,the work done in this paper also promoted the research and development of GPU general computing in the field of digital image encryption.