Research on Image Steganography Information Detection Based on Support Vector Machine

Abstract

With the rapid development of the internet of things and cloud computing, users can instantly transmit a large amount of data to various fields, with the development of communication technology providing convenience for people's life, information security is becoming more and more important. Therefore, it is of great significance to study the technology of image hiding information detection. This paper mainly uses the support vector machine learning algorithm to detect the hidden information of the image, based on a standard image library, randomly selecting images for embedding secret information. According to the bit-plane correlation and the gradient energy change of a single bit-plane after encryption of an image LSB matching algorithm, gradient energy change is selected as characteristic change, and the gradient energy change is innovatively applied to a support vector machine classifier algorithm, And has very good detection effect and good stability on the dense image with the embedding rate of more than 40 percent. Keywords-Support Vector Machine; Information Detection; LSB Matching; Steganography