Lossless Compression-Based Steganalysis of LSB Embedded Images

Abstract

A new method of steganalysis for images with embedded messages is presented. We consider two embedding methods: least significant bit (LSB) replacement and plusmn1 LSB embedding. Our method uses lossless image compression to generate statistics that are fed into a support vector machine classifier. We compare results against the pairs method, one of the best existing LSB steganalysis methods. Both our method and pairs performs well for LSB replacement. However, while pairs cannot detect plusmn1 LSB embedding at all, our method can.