System and method for image steganalysis

Abstract

The present invention provides a system and method for image steganalysis, relating to the fields of network transmission, information security, digital image processing, machine learning and the like. The method comprises a step of establishing an original image database, generating a steganographic image database for the image hidden information in the original image database, extracting the characteristics of the original image database and the images in steganographic images, and generating multiple sample spaces, a step of carrying out sample training on the multiple sample spaces, generating multiple classifiers, combining the multiple classifiers to generate a new classifier through a voting mode, and carrying out steganalysis on the image uploaded by the user through the new classifier. According to the system and the method, different steganographic methods can be selected to carry out steganography on the image, the characteristic of an image set is extracted and is converted into vector data, and different characteristic extracting methods can be employed at the same time.

Description

technical field [0001] The invention relates to the fields of network transmission, information security, digital image processing, machine learning, etc., and in particular to a system and method for image steganalysis. Background technique [0002] Steganography is a technique and science about information hiding, that is, to prevent anyone other than the planned recipient from knowing the information transmission event (not just the content of the information), so that it can effectively ensure that information is transmitted on the network. Steganalysis refers to detecting and judging whether there is secret information from the observed data in the case of known or unknown embedding algorithms, analyzing the size of the data volume and the location of data embedding, and finally cracking the embedded content. process. [0003] In recent years, the field of steganography has become the focus of information security, because every Web site relies on multimedia, such as a...

AI Technical Summary

Problems solved by technology

[0007] The above general blind detection methods for information hiding are not blind detection in the true sense, but "half blind" detection under certain conditions. However, in actual detection, when analysts detect images, the objective situation often cannot meet this condition

[0008] In conclusion, although the existing detection methods have good detection performance for some hidden methods, there are still some key issues worth exploring, such as: low embedding rate detection problem, image source mismatch problem, steganographic image detection algorithm applicable Problems of detection range, applicability of steganalysis methods, etc.

Images