作者: Hassan Farsi , Amir Shahi
DOI: 10.1080/02533839.2014.929711
关键词:
摘要: The aim of digital image steganalysis is to detect hidden information (which can be a message or an image) in steganographic image. An ideal steganography method encrypts the such that it cannot easily detected. Currently, wide variety different techniques are being used; therefore, more advanced methods needed images coded by techniques. A typical technique consists two parts: (1) feature extraction and (2) classification. In this paper, new based on Markov chain process proposed. proposed technique, after features, non-linear classifier named support vector machine applied classify clean encrypted images. Analysis variance used reduce dimensions features. performance compared against subtractive DCT coefficient adjacency matrix (SDAM) subtrac...