To solve the problems of existing image steganography analysis methods,such as poor generalizability and poor detection accuracy,a big data analysis method for image steganography was proposed.Firstly,according to the characteristics of the adaptive steganography algorithms,the high complexity region of the image to be tested was intercepted as the core analysis image.Secondly,the image matching was completed using database to get the steganographic difference features between the matched image and the core analysis image.Finally,the features were input into the convolutional neural network model to complete the detection and classification of the steganography algorithm.The experimental results show that the average accuracy of this method for the detection of six steganography algorithms reaches 93.98％under the condition of successful matching by default.The method supports images in both the spatial and the frequency domain,and possesses strong versatility.
万方数据
维普
