Equal-mode vector decomposition image encryption analysis method based on novel full convolutional network

Abstract

The invention provides an equal-mode vector decomposition image encryption analysis method based on a novel full convolutional network. The method comprises four parts of an encryption system based onequal-modulus vector decomposition, a network model of encryption analysis and network training and encryption system analysis. A designed encryption analysis network model is trained by inputting aplaintext-ciphertext pair, and then a ciphertext image is input through the trained encryption analysis network model, so that an attack analysis result, namely a recovered high-quality plaintext image, can be obtained. Compared with a traditional attack method, the provided encryption analysis method has the advantages that effective encryption analysis can be realized without knowing an encryption secret key or a private key, other encryption system parameters and the like, and a high-quality plaintext image can be recovered; the provided deep learning method is short in training time, and the training speed is increased by 7 times compared with a traditional method; the method provided by the invention has good generalization ability, and one image library can be adopted for training while the other image library can be adopted for successful testing; finally, the method also has good robustness for noise and clipping in transmission.

Description

technical field [0001] The invention relates to the technical fields of information security and information optics, in particular to an image encryption analysis method. Background technique [0002] With the advent of the information age, information security has received more and more attention. Images often provide rich information, so image encryption becomes a crucial issue. In 2015, Cai proposed an image encryption method based on equal modulus separation (EMD), which decomposes a two-dimensional vector into two two-dimensional vectors, and provides a secure one-way trapdoor function for image encryption, so it is widely used in grayscale image encryption. As an effective method of feature extraction, deep learning has also been used in image encryption analysis, but currently only encryption systems based on double random phase encoding have been broken by deep learning methods. For asymmetric encryption methods with higher encryption security such as The image en...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the above-mentioned problem that the traditional deep learning network cannot attack and analyze image encryption based on equal modulus vector decomposition, the present invention proposes a novel full convolutional network-based equal modulus vector decomposition image encryption analysis method

Images