混沌线程池与GPU优化的批量图像加密算法

数据量大且冗余度高是数字图像显著的特征,这对大批量图像快速实时加密提出了挑战。为了解决此问题,基于Lorenz混沌加密技术,设计了一种采用线程池与图形处理器(graphics processing unit,GPU)组合优化的批量图像加密算法。该算法通过线程池改进图像的读写,并进行图像镜像变换;利用Lorenz混沌系统生成加密序列,结合图像分块混沌序列进行加密;然后对批量图像数据进行打包,通过GPU进行大批量的异步计算;最后重组图像矩阵得到批量加密图像。实验测试表明,该算法能够有效抵御常见的攻击手段,经过性能优化后的批量数字图像加密算法,可以保证图像安全性;同时,在批量图像读取速率和加解密处理效率方面有显著的提高。

Batch Image Encryption Algorithm of the Optimization of Chaotic Thread Pool and GPU

Large data volume and high redundancy are the prominent features of digital images, which are challenges for the fast and real-time encryption of batch images. To solve this problem, a batch image encryption algorithm was designed based on Lorenz chaotic encryption technology, which was optimized by a combination of the thread pool and GPU. The algorithm applies thread pool to improve image read-write, and performs image transformation; It applies Lorenz chaotic system to generate the encryption sequence, and encrypts the image by combining the chaotic sequence of image blocks. Then the image data is packaged and the asynchronous calculation is carried out by GPU. Finally, the image matrix is reassembled to obtain batch-encrypted images. The experimental results show that the algorithm can effectively resist common attacks, and the optimized batch digital image encryption can ensure image security. At the same time, the efficiency of read-write and the processing of encryption and decryption for batch images was significantly improved.