Design of Lossless Compression Encryption Algorithm for Dynamic Image under Deep Learning
Image compression is aimed at reducing redundant information in image data,making it more efficient and secure for storage and transmission.Since sensitive information exists in dynamic images,it is easy to cause dam-age or loss of data during transmission.In order to enhance the security of lossless compression encryption of dynamic images,a method for lossless compression encryption of dynamic images was proposed based on deep learning.Firstly,the distribution rules of redundant information in dynamic images were analyzed,and then the prediction tem-plates were constructed.Differential Pulse Code Modulation(DPCM)was used for lossless encryption of transformed data.Moreover,the deep learning algorithm was used to extract features from dynamic images.Furthermore,individual features of dynamic images were used to form a key.After that,it was applied to dynamic image encryption.Finally,the lossless compression encryption of dynamic images was achieved.Experimental results show that the proposed method has a high peak signal-to-noise ratio,and the compression ratio is more than 1.7％,indicating its superiority in dynamic image lossless compression,high encryption effect as well as security.
Deep learningDynamic imageLossless compressionEncryption
NETL
NSTL
万方数据
