Steganalysis Based on JPEG Binary Compressed Stream
Hongtao Lu · 2009
This paper proposes a method of steganalysis based on JPEG binary compressed stream, which is different from the steganalysis based on Discrete Cosine Transform(DCT) frequency domain and image spatial domain. It utilizes the changes of binary compressed data stream caused by the hidden messages, and uses Multiple Embedding Method(MEM) to draw a feature vector, which is used to train the One Class Support Vector Machine(OC-SVM). OC-SVM get high classification rate towards JSteg and Outguess hiding methods in the experiment.