Abstract
dc:descriptionNowadays, steganography is the main mean of illegal secret communication. Therefore, the need of detecting steganographic content and especially stego images is becoming more compulsory. However, steganalysis is a very difficult task and its success depends on many factors, like the presence of the cover medium, evidence of the utilized steganographic algorithm etc. Early steganalysis methods deploy statistical attacks on stego images while more recent ones use deep learning techniques. The latter ones mainly utilize convolutional neural networks and show promising results. This dissertation deals with issues related to steganalysis and in particular to image steganalysis. Βasic concepts of image steganalysis along with a taxonomy for classification of the different steganalysis methods used by a digital forensic examiner are presented. Moreover, a detailed overview of state-of-the-art methods proposed in literature is given. The research focuses in two major research questions i.e. the proposal of a novel convolutional neural network, and afterwards its utilization as a feature extractor. The proposed method initially utilized a dilated convolutional neural network - KarNet - to identify stego images from two different steganographic algorithms i.e. Spatial-Universal Wavelet Relative Distortion (S-UNIWARD) and Wavelet Obtained Weights (WOW). The proposed convolutional neural network was compared against other state-of-the-art deep learning techniques and it outperforms them.
Degree
thesis:*- Grantor dc:publisher
- University of the Aegean
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Karampidis, Konstantinos
- Καραμπίδης, Κωνσταντίνος
Subjects
dc:subject × 15- Στεγανάλυση
- Βαθιά μάθηση
- Ψηφιακή εγκληματολογία
- Μηχανική μάθηση
- Συνελικτικά νευρωνικά δίκτυα
- Steganalysis
- Deep learning
- Digital forensics
- Machine learning
- Convolutional neural networks
- Pattern recognition
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
dc:rights- Statement dc:rights
-
- BY_NC_SA
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/47692