Image steganography analysis method and device

Abstract

The invention discloses an image steganography analysis method and device. The method comprises the following steps: (1) acquiring a gallery; (2) extracting steganography characteristics; (3) acquiring fern characteristics; (5) acquiring probability distribution, and establishing a probability control table; (6) processing images to be detected, acquiring corresponding occurrence probability of each fen value of each image to be detected in steganography images and corresponding occurrence probability of each fen value of each image to be detected in carrier images, and comparing the two corresponding probabilities, so as to obtain whether the corresponding image is a steganography image or not. According to the method and the device, the probability distribution of the steganography images and no-steganography images is obtained through training a database, and whether the images to be detected are the steganography images or not are judged through processing the images to be detected, so that the method has universality and can be applicable to different steganography ways.

Description

technical field [0001] The invention belongs to the technical field of information security, and relates to a steganalysis method, in particular to an image steganalysis method and a device thereof. Background technique [0002] With the development of computer and Internet technology, information security issues have become increasingly prominent. In important fields such as national security agencies, intelligence and security departments, and national defense forces, how to effectively ensure the confidentiality, integrity, availability, and controllability of digital information is an important issue to be solved in the current information security field. In the army command information system, all kinds of texts and messages transmitted are of high confidentiality level, once intercepted or destroyed by the enemy, it will cause immeasurable losses. The traditional communication transmission method uses cryptography to convert text messages into ciphertext, but the bigg...

AI Technical Summary

Problems solved by technology

[0007] The integrated classifier for steganalysis proposed by Kodovsky et al. is currently the preferred classifier for processing high-dimensional steganalysis features. The final result is obtained by voting on the detection results of several base classifiers. Although the training complexity is reduced, the classification accuracy and prediction speed to be improved

Based on hypothesis testing theory, Cogranne et al. theoretically proved the effectiveness of steganalysis for integrated classification, but did not fully consider the difference in prediction performance between different base classifiers, and with the increase of the number of base classifiers, not only the prediction speed decreased significantly , the storage space required will also increase rapidly

Images