Digital Media Steganography: Principles, Algorithms, and Advances

Cover
Digital Media Steganography:
Principles, Algorithms, and Advances
Copyright
Contents
List of contributors
About the editor
Preface
Acknowledgments
1 Introduction to digital image steganography
	1.1 Introduction
	1.2 Applications of steganography
	1.3 Challenges facing steganography
	1.4 Steganographic approaches
		1.4.1 Spread spectrum approaches
		1.4.2 Spatial domain approaches
			1.4.2.1 Gray level modification
			1.4.2.2 Pixel value differencing (PVD)
			1.4.2.3 Least significant bit substitution (LSB)
			1.4.2.4 Exploiting modification direction (EMD)
			1.4.2.5 Quantization-based approaches
			1.4.2.6 Multiple bit-planes-based approaches
		1.4.3 Adaptive-based approaches
		1.4.4 Transform domain approaches
	1.5 Performance evaluation
		1.5.1 Payload capacity
		1.5.2 Visual stego image quality analysis
		1.5.3 Security analysis
			1.5.3.1 Pixel difference histogram analysis
			1.5.3.2 Universal steganalysis
			1.5.3.3 Regular and singular steganalysis
	1.6 Conclusion
	References
2 A color image steganography method based on ADPVD and HOG techniques
	2.1 Introduction
	2.2 Review of the ADPVD method
	2.3 The pixel-based adaptive directional PVD steganography
		2.3.1 Histogram of oriented gradients
		2.3.2 Pixel-of-interest (POI)
		2.3.3 Embedding algorithm
		2.3.4 Extraction algorithm
	2.4 Results and discussion
		2.4.1 Embedding direction analysis using HOG
		2.4.2 Embedding direction analysis using POI
		2.4.3 Impact of threshold value on POI
		2.4.4 Impact of threshold on capacity and visual quality
		2.4.5 Visual quality analysis
		2.4.6 Comparison with other adaptive PVD-based methods
		2.4.7 Comparison with color image-based methods
		2.4.8 Comparison with edge-based methods
		2.4.9 Security against pixel difference histogram analysis
		2.4.10 Security against statistical RS-steganalysis
	2.5 Conclusion
	References
3 An improved method for high hiding capacity based on LSB and PVD
	3.1 Introduction
	3.2 Related work
		3.2.1 Pixel value differencing (PVD) steganography [13]
			3.2.1.1 The PVD embedding procedure
			3.2.1.2 The PVD extraction steps
			3.2.1.3 Illustration of the PVD method
		3.2.2 Khodaei et al.'s method [20]
			3.2.2.1 An illustration of incorrect data extraction in Khodaei et al.'s method
		3.2.3 Jung's method [15]
			3.2.3.1 Embedding algorithm
			3.2.3.2 Extraction algorithm
			3.2.3.3 FOBP in Jung's method
			3.2.3.4 Extraction problem in Jung's method
	3.3 The proposed method
		3.3.1 Embedding procedure
			Case 1: Pixel shifting process for overflow condition
			Case 2: Pixel shifting process for underflow condition
		3.3.2 Extraction procedure
		3.3.3 Example of the proposed method
			3.3.3.1 Embedding side
			3.3.3.2 Extraction side
	3.4 Results and discussion
		3.4.1 Analysis of PSNR, capacity, BPP, FOBP, and SSIM
		3.4.2 Security check using RS analysis
		3.4.3 Security check using Pixel Difference Histogram (PDH) analysis
	3.5 Conclusion
	References
4 An efficient image steganography method using multiobjective differential evolution
	4.1 Introduction
	4.2 Literature review
	4.3 Background
		4.3.1 LSB substitution method
		4.3.2 Differential evolution
	4.4 The proposed method
		4.4.1 Embedding process
		4.4.2 Extraction process
	4.5 Experimental results
		4.5.1 Peak signal-to-noise ratio
		4.5.2 Structural similarity index measure
		4.5.3 Bit error rate
	4.6 Conclusion
	References
5 Image steganography using add-sub based QVD and side match
	5.1 Introduction
	5.2 Proposed ASQVD+SM technique
		5.2.1 The embedding procedure
		5.2.2 Extraction procedure
		5.2.3 Example of embedding and extraction
	5.3 Experimental analysis
	5.4 Conclusion
	References
6 A high-capacity invertible steganography method for stereo image
	6.1 Introduction
	6.2 Preliminaries
		6.2.1 Discrete cosine transforms (DCT) and quantized DCT (QDCT)
		6.2.2 Yang and Chen's method
	6.3 The proposed method
		6.3.1 Generation of the embedding direction histogram (EDH)
		6.3.2 Stereo image embedding algorithm
			6.3.2.1 Similar block searching
			6.3.2.2 Based-2-D histogram shifting with EDH data embedding
			6.3.2.3 Example of embedding
		6.3.3 Information extracting and stereo image recovering algorithm
		6.3.4 Evaluation metrics
	6.4 Experimental results
	6.5 Conclusion
	Acknowledgment
	References
7 An adaptive and clustering-based steganographic method: OSteg
	7.1 Introduction
	7.2 Related works
	7.3 OSteg embedding
		7.3.1 Preparation
		7.3.2 Otsu clustering
		7.3.3 Pretreatment: fake embedding
		7.3.4 Scrambling selection: Ikeda system
		7.3.5 Secret shared key and key space
		7.3.6 Effective embedding
	7.4 Experimental results and discussion
	7.5 Conclusion
	Acknowledgments
	References
8 A steganography method based on decomposition of the Catalan numbers
	8.1 Introduction
	8.2 Related works
	8.3 Decomposition of Catalan numbers
	8.4 Implementation of the proposed method
		Module for embedded data
		Module for extract data
	8.5 Steganalysis and security testing
		Security analysis of stego key
		Steganalysis of the proposed method
	8.6 Conclusion
	References
9 A steganography approach for hiding privacy in video surveillance systems
	9.1 Introduction
	9.2 Related works
	9.3 Hiding privacy information using video compression concept
		9.3.1 Background model generator
		9.3.2 Deidentification private details
		9.3.3 H.264 compression preprocessing
		9.3.4 The proposed quantization hiding technique
		9.3.5 The extraction module
	9.4 Experimental results
		9.4.1 Data payload
		9.4.2 Invisibility performance
	Conclusion
	References
10 Reversible steganography techniques: A survey
	10.1 Introduction
		10.1.1 Reversible Steganography Scheme (RSS)
		10.1.2 Measurements of RSS
		10.1.3 Categories of RSS
	10.2 Difference Expansion (DE) schemes
		10.2.1 Embedding procedure of Tian's method
		10.2.2 Extraction procedure of Tian's method
		10.2.3 Embedding procedure of Alattar's method
		10.2.4 Extraction procedure of Alattar's method
		10.2.5 Recovery procedure of Alattar's method
	10.3 Histogram-Shifting (HS) schemes
		10.3.1 Embedding procedure of HS
		10.3.2 Extraction and recovery procedures of HS
		10.3.3 Extra information of HS
		10.3.4 Experimental results of HS
	10.4 Pixel-Value-Ordering (PVO) schemes
		10.4.1 Embedding procedure of PVO
		10.4.2 Embedding procedure of IPVO
		10.4.3 Experimental results of PVO-based schemes
	10.5 Dual-image-based schemes
		10.5.1 Center-folding strategy
		10.5.2 Experimental results of dual-based RSS
	10.6 Interpolation-based schemes
		10.6.1 Embedding procedure of NMI
		10.6.2 Extraction procedure of NMI
		10.6.3 Comparison results
	10.7 Conclusion
	Acknowledgments
	References
11 Quantum steganography
	11.1 Introduction
		11.1.1 The idea of steganography
		11.1.2 Quantum error-correcting codes
	11.2 Goals and tools of quantum steganography
	11.3 Quantum steganography with depolarizing noise
		11.3.1 The depolarizing channel
		11.3.2 A local steganographic encoding
		11.3.3 Key usage
		11.3.4 Weaknesses of the local encoding
	11.4 Steganographic encoding in error syndromes
		11.4.1 The encoding and decoding procedure
		11.4.2 Communication and key usage rates
	11.5 Encoding in the binary symmetric channel
	11.6 Encoding in the 5-qubit "perfect" code
		11.6.1 Encoding with one-qubit errors
		11.6.2 Two error encodings
		11.6.3 Rate of secret qubit transmission
		11.6.4 Comparison to encoding across blocks
			Steganographic communication rate
			Key usage rate
	11.7 Secrecy and security
		11.7.1 Diamond norm distance for the binary symmetric channel
		11.7.2 Diamond norm distance for the depolarizing channel
		11.7.3 Conditions for secrecy
		11.7.4 Secret key vs. shared entanglement
	11.8 Asymptotic rates in the noiseless case
		11.8.1 Direct coding theorem (achievability)
			The binary symmetric channel
			The depolarizing channel
			Random unitary channels
			General channels
			Secret key consumption
		11.8.2 Converse theorem (upper bound)
			Upper bound on steganographic rate
	11.9 Asymptotic rates in the noisy case
		11.9.1 Direct coding in the noisy case
			Achievable rate for the BSC
			Secret key consumption
			Depolarizing channel
			General channels
		11.9.2 Converse theorem in the noisy case
			Upper bound on steganographic rate
	11.10 Discussion and future directions
	11.11 Conclusion
	Acknowledgments
	References
12 Digital media steganalysis
	12.1 Introduction
	12.2 Image steganalysis
		12.2.1 Signature steganalysis
		12.2.2 Statistical steganalysis
		12.2.3 Deep learning applied to steganalysis of digital images
		12.2.4 Summary and perspectives
	12.3 Audio steganalysis
		12.3.1 Methods
			12.3.1.1 Noncompressed audio formats
			12.3.1.2 Compressed audio formats
			12.3.1.3 Modern audio steganalysis
		12.3.2 Summary and perspectives
	12.4 Video steganalysis
		12.4.1 General context
		12.4.2 Previous methods
		12.4.3 Recent method
		12.4.4 Summary and perspectives
	12.5 Text steganalysis
		12.5.1 Methods
			12.5.1.1 Statistical algorithms
			12.5.1.2 Modern text steganalysis
		12.5.2 Summary and perspectives
	12.6 Conclusion
	References
13 Unsupervised steganographer identification via clustering and outlier detection
	13.1 Introduction
	13.2 Primary concepts and techniques
		13.2.1 JPEG compression
		13.2.2 JPEG steganalysis features
			13.2.2.1 PEV-274 features
			13.2.2.2 LI-250 features
		13.2.3 Batch steganography and pooled steganalysis
		13.2.4 Agglomerative clustering
		13.2.5 Local outlier factor
		13.2.6 Maximum mean discrepancy
	13.3 General frameworks
		13.3.1 Clustering-based detection
		13.3.2 Outlier-based detection
		13.3.3 Performance evaluation and analysis
			13.3.3.1 Clustering-based detection
			13.3.3.2 Outlier-based detection
	13.4 Ensemble and dimensionality reduction
		13.4.1 Clustering ensemble
		13.4.2 Dimensionality reduction
			13.4.2.1 Feature selection
			13.4.2.2 Feature projection
	13.5 Conclusion
	Acknowledgment
	References
14 Deep learning in steganography and steganalysis
	14.1 Introduction
	14.2 The building blocks of a deep neuronal network
		14.2.1 Global view of a Convolutional Neural Network
		14.2.2 The preprocessing module
		14.2.3 The convolution module
		14.2.4 The classification module
	14.3 The different networks used over the period 2015-2018
		14.3.1 The spatial steganalysis Not-Side-Channel-Aware (Not-SCA)
		14.3.2 The spatial steganalysis Side-Channel-Informed (SCA)
		14.3.3 The JPEG steganalysis
		14.3.4 Discussion about the Mismatch phenomenon scenario
	14.4 Steganography by deep learning
		14.4.1 The family by synthesis
		14.4.2 The family by generation of the modifications probability map
		14.4.3 The family by adversarial-embedding iterated (approaches misleading a discriminant)
		14.4.4 The family by 3-player game
	14.5 Conclusion
	References
Index
Back Cover