This thesis presents the development of novel digital image watermarking technique for the shielding of intellectual property rights of digital images. Because of their digital nature, multimedia data can be duplicated, modified, and transformed. In this context, it is essential to develop a watermarking –based technique for copyright protection and authentication of digital images. The major problem of watermarking technique is how to achieve an optimal-tradeoff between robustness and imperceptibility. A robust digital image watermarking technique has been developed based on cascaded Discrete Orthonormal Stockwell Transform (DOST), Discrete Wavelet Transform (DWT), and Singular Value Decomposition (SVD) using private key obtained from the implementation of a Quantum Key Distribution (QKD) scheme and optimal scaling factor selection using Particle Swarm Optimization (PSO) algorithm. The colour images acquired are standard colour Lena, Pepper, Mandril, and Nepal Telecom logo. The colour images were decomposed into the respective three colour channels, of Red (R), Green (G), and Blue (B). The developed technique is achieved by applying DOST scheme on each channel of the cover and watermark images to obtain DOST transformed coefficients, these are then fed to the DWT for further transformation by decomposing the coefficients up to third level for complete representation and interpretation of the images in order to produce robust and scalable images. The images are modified by applying SVD to improve the security. A key length of 64-bits generated from the implementation of QKD scheme is used to encrypt the transformed watermark to further enhance the security of the technique. The watermark is reshuffled by affine transform by redistributing the pixel values to different locations using four 8-bit keys for it to be robust to attacks. The unintelligible transformed image by affine transform is divided into 2 pixels x 2 pixels blocks and each of the block is encrypted using XOR operation by four 8-bit keys. The optimal scaling factor obtained is used to embed and extract the watermark. The performance of the scheme is evaluated using Peak Signal to Noise Ratio (PSNR), Normalized Correlation Coefficient (NCC), and Structural Similarity Index Measure (SSIM) as performance metrics for imperceptibility and robustness. The results of the PSNR obtained for accessing the fidelity of the watermarked image when subjected to Gaussian, salt and pepper, speckle, rotated at 45°, 5°, and 90°, and crop attacks are 50.2752dB, 47.7293dB, 45.6404dB, 46.8626dB, 44.6045dB, 47.9442dB and 46.4067dB respectively, indicating that perceptual quality has been improved. The NCC values are 1, 1, 1, 0.9997, 1, 1 and 1, signifying high resistance to attacks. The SSIM values are 1, 1, 1, 0.99998, 0.9996, 0.9999 and 0.9987, respectively. The developed scheme with the QKD scheme delivered an imperceptibility improvement when compared with ITU-R standard value of 35dB for watermarking technique by 30%, 27%, 23%, 25%, 22%, 27% and 25% respectively. Comparison of the developed algorithm with the work of Bajracharya & Koju (2017) in terms of perceptual quality of the watermarked image showed an average PSNR value of 64.3580dB as against 48.1819, representing about 25% improvement of imperceptibility. The NCC values of the work of Bajracharya & Koju (2017) when subjected to the following attacks: Gaussian, salt and pepper, speckle, rotated at 45°, 5°, and 90°, and crop attacks are 1, 0.9996, 0.9999, 0.9999, 1 and 1. This demonstrates that the developed scheme gave an average improvement of about 14% robustness to attacks when compared with the work of Bajracharya & Koju (2017).
TABLE OF CONTENTS
TITLE PAGE i
LIST OF FIGURES xii
LIST OF TABLES xxi
LIST OF ABBREVIATION xxii
CHAPTER ONE: INTRODUCTION
1.1 Background of Research 1
1.2 Motivation 10
1.3 1.3 Significance of Research 12
1.4 Statement of Problem 13
1.5 Aim and Objectives 14
1.6 Scope of the Research 15
1.7 Thesis Organization 15
CHAPTER TWO: LITERATURE REVIEW
2.1 Introduction 16
2.2 Review of Fundamental Concepts 16
2.2.1 Digital Image Watermarking 16
2.2.3 Characteristics of Digital Image Watermarking 17
18.104.22.168 Imperceptibility 17
22.214.171.124 Capacity 18
126.96.36.199 Security 18
188.8.131.52 Robustness 19
184.108.40.206 Computational Complexity 20
220.127.116.11 Verifiability 20
2.2.4 Applications of Digital Image Watermarking 21
2.2.5 Digital image watermarking techniques 22
18.104.22.168 Discrete Wavelet Transform (DWT) 25
22.214.171.124 Discrete Fourier Transform (DFT) 28
126.96.36.199 Singular Value Decomposition (SVD) 28
188.8.131.52 Stockwell Transform (ST) 30
184.108.40.206 Discrete Orthonormal Stockwell Transform (DOST) 32
2.2.6 Comparison between Digital Image Watermarking Techniques 36
2.2.7 Attack on Digital Image Watermarking 39
220.127.116.11 Signal Processing Attacks 40
18.104.22.168 Geometric Attacks 40
2.2.8 Performance Evaluation Metrics 41
2.2.9 Cryptography 42
22.214.171.124 Challenges of Classical Key Distribution 44
126.96.36.199 Quantum Cryptography 45
188.8.131.52 Quantum Key Distribution (QKD) 47
2.2.10 Bennet and Brassard 1984 (BB84) Protocol 50
184.108.40.206 Steps of QKD Processes Based on the BB84 Protocol 52
220.127.116.11 Suitable Selection of Scaling Factor 55
2.2.11 Particle Swarm Optimization 55
2.3 Review of Similar Works 57
CHAPTER THREE : MATERIALS AND METHODS
3.1 Introduction 73
3.2 Materials 73
3.3 Important Assumptions 73
3.4 Methodology 74
3.4.1 Development of a Digital Image Watermarking Technique 76
3.4.2 Application of 2D DOST on the Cover and Watermark images 77
CHAPTER FOUR: RESULTS AND DISCUSSION
4.1 Introduction 97
4.2 Decomposition of the Cover and Watermark Images 97
4.3 Determination of Scaling Factor 98
4.4 Definition of Different Scenarios 98
4.4.1 Robustness and Imperceptibility Results for Case I 99
4.4.2 Robustness and Imperceptibility Results for Case II 100
4.4.3 Robustness and Imperceptibility Results for Case III 101
4.4.4 Robustness and Imperceptibility Results for Case IV 102
4.4.5 Robustness and Imperceptibility Results for Case V 104
4.4.6 Robustness and Imperceptibility Results for Case VI 106
4.4.7 Robustness and Imperceptibility Results for Case VII 108
4.5 Other Scenarios Considered for Comparison 111
4.5.1 Robustness and Imperceptibility Results for Case VIII 111
4.5.2 Robustness and Imperceptibility Results for Case IX 112
4.5.3 Robustness and Imperceptibility Results for Case X 113
4.5.4 Robustness and Imperceptibility Results for Case XI 114
4.6 Determination of Optimal Scale Factor Based on Individual Attack Scenarios using PSO 116
4.7 Definition of Different Scenarios 117
4.7.1 Robustness and Imperceptibility Results for Case I 118
4.7.2 Robustness and Imperceptibility Results for Case II 119
4.7.3 Robustness and Imperceptibility Results for Case III 120
4.7.4 Robustness and Imperceptibility Results for Case IV 121
4.7.5 Robustness and Imperceptibility Results for Case V 122
4.7.5 Robustness and Imperceptibility Results for Case VI 123
4.8 Definition of Different Scenarios 125
4.8.1 Robustness and Imperceptibility Results for Case I 126
4.8.2 Robustness and Imperceptibility Results for Case II 127
4.8.3 Robustness and Imperceptibility Results for Case III 129
4.8.4 Robustness and Imperceptibility Results for Case IV 130
6.8.5 Robustness and Imperceptibility Results for Case V 131
4.8.6 Robustness and Imperceptibility Results for Case VI 132
4.8.7 Robustness and Imperceptibility Results for Case VII 133
4.9 Definition of Different Scenarios 135
4.9.1 Robustness and Imperceptibility Results for Case I 136
4.9.2 Robustness and Imperceptibility Results for Case II 138
4.9.3 Robustness and Imperceptibility Results for Case III 139
4.9.4 Robustness and Imperceptibility Results for Case IV 141
4.9.5 Robustness and Imperceptibility Results for Case V 143
4.9.6 Robustness and Imperceptibility Results for Case VI 144
4.9.6 Robustness and Imperceptibility Results for Case VII 145
4.10 Performance Evaluation of the Developed Algorithm 148
4.11 Comparison of Results 149
4.11.1 Comparison of PSNR, NCC and SSIM Results of Various Defined Scenarios under No Attack 149
4.11.2 Comparison of PSNR, NCC and SSIM Results of Various Defined Scenarios under Gaussian Attack 152
4.11.3 Comparison of PSNR, NCC and SSIM Results of Various Defined Scenarios under Salt and Pepper Attack 154
4.11.4 Comparison of PSNR, NCC and SSIM Results of Various Defined Scenarios under Speckle Attack 157
4.12 Validation 159
CHAPTER FIVE: CONCLUSION AND RECOMMENDATION
5.1 Summary 163
5.3 Conclusion 164
5.4 Significant Contributions 167
5.5 Limitations 167
5.6 Recommendations for future work 168
APPENDIX A 177
Complete M File for Host Image Decomposition Using DOST 177
APPENDIX B 179
Complete M File for Binary bits Generation 179
APPENDIX C 180
Complete M File for Generating and Sending of Quantum bits to Receiver 180
APPENDIX D 181
Complete M File for Receiving Quantum bits from Sender 181
APPENDIX E 183
Complete M File for performing the various phases of Error Detection and Correction 183
APPENDIX F 185
Complete M File for performing Error Reconciliation 185
APPENDIX G 188
Complete M File for performing Error Reconciliation process on the key obtained 188
APPENDIX H 191
Complete M File for Watermark Encryption with Generated Key 191
APPENDIX I 192
Complete M File for Application of transformation Technique 192
APPENDIX J 193
Complete M File for Performing the Fitness Function 193
APPENDIX K 194
Complete M File for Decrypting the watermark image 194
APPENDIX L 195
Complete M File for using PSO to obtain optimal Scale factor based on Fitness Function 195
APPENDIX M 196
Complete M File for Watermarking Process 196
APPENDIX N 198
Complete M File for Host Image Recovering Using Inverse DOST 198
APPENDIX O 200
Complete M File for Application of Inverse Transformation 200
APPENDIX P 201
Complete M File for Generating the watermarking simulation user interface 201
APPENDIX Q 203
Summary of results obtained for the various Attack scenarios 203
APPENDIX R 204
Summary of results obtained of Optimal Scaling Factor under different Attack Scenarios 204
APPENDIX S 206
Comparison of the random bits generated by Alice, Alice randomly chooses polarization basis, transmitted Qubits by Alice and Bob measures Qubits 206
1.1 Background of Research
The rising demand for the production, storage and transmission of multimedia contents (such as image, audio, video and text) over secured and unsecured communication media in recent years poses a lot of security and privacy concerns to both the sender and receiver. The use of these multimedia contents (image, audio, video and text) is rapidly increasing with the high growth and widespread use of the Internet and information technology. Due to this fact, tampering with and illegal distribution of digital contents is inevitable and as such it becomes imperative to devise mechanisms to protect the copyright of such media. It has been established that present copyright laws are insufficient for addressing the security of digital contents (Chandramouli et al., 2002). Furthermore, simple transfer and manipulation of digital data also institutes a real menace for information inventors, and copyright owners want to be recompensated every period their work is used. In addition, they want to be certain that their work is not deployed in an illegitimate means (for instance modified without their consent).
The introduction of Internet has resulted in numerous opportunities for the creation and transfer of contents (electronic advertising, web publishing, digital repositories and libraries, real-time video and audio delivery, etc.) in digital form (Chandramouli et al., 2002). A pertinent issue that arises in these applications is the protection of the rights of all participants, such as copyright enforcement and content verification because the present copyright acts are insufficient in sharing digital messages (Chandramouli et al., 2002). One solution to this threat would be to curb access to the data using encryption technique (Arnold et al., 2003), even though it does not necessarily offer total protection. Once the encoded data are decoded,
they can be easily circulated or manipulated (Arnold et al., 2003), this has led to the fascination of researchers in developing copy deterrence and protection mechanisms to overcome illegal manipulation of digital data (Kavitha & Shan, 2016).
Several mechanisms have been proposed for the protection of multimedia contents based on data hiding techniques. These techniques are as follows (Harish et al., 2013):
A comprehensive review of the development of data hiding can be found in (Tanaka, Nakamura & Matsui, 1990).
Cryptography is not particularly focused on concealing the presence of data, but is also regarded as encryption (Challita & Farhat, 2011). Cryptography can be defined as the study of mathematical systems for solving two types of challenges, privacy and authentication (Diffie & Hellman, 1976). A privacy scheme precludes the extraction of information by unauthorized parties from data transferred over public channel, hence ensuring the sent message is being read only by the designated receiver (Diffie & Hellman, 1976). An authentication scheme is to forestall an unauthorized injection of messages into public channel, ascertaining the recipient of a message of the legitimacy of its sender (Diffie & Hellman, 1976).
Steganography means the study of techniques for concealing the existence of a secondary message in the presence of a primary message (Arnold et al., 2003), i.e. by inserting a private message in a cover image (Challita & Farhat, 2011). Where the carrier signal depicts the primary message and the secondary message is the payload signal or payload message. Steganography is a mechanism that is used for offering confidentiality and deniability, both of
which can be satisfied exclusively through cryptographic ways (Arnold et al., 2003). In steganography, messages which are secreted has no connection with the cover channel or image and the condition for steganography is that no intuition should arise that a channel is conveying any concealed data (Challita & Farhat, 2011). The aim of steganography is to have covert communication between two parties, that is, presence of the communication is unknown to a potential assailant, and only a fruitful attack can notice the existence of this conversation.
One of such mechanisms that have been attracting major interest is digital watermarking (Averkiou, 2002). Watermarking has provides one of the best solutions among steganography and cryptography mechanisms (Harish et al., 2013). Watermarking has shown resilient property that overcomes covert communication as in steganography (Harish et al., 2013). The interest in watermarking actually started in 1990 with the development of the multimedia systems and the necessity of transferring data over the internet (Araghi et al., 2016). Digital watermarking is a technique that is used to preclude duplicating or to shield digital data by invisibly hiding lawful marked message into the original data (Lai et al., 2013). Digital watermarking can also be defined as the act of concealing information connected to a digital signal (which could be a video, song and image) inside the signal itself (Bajracharya & Koju, 2017). Watermarking attempts to hide a message interrelated to the actual content of the digital signal (Bajracharya & Koju, 2017). Such information is embedded for various reasons such as (Nin & Ricciardi, 2013):
1. Copyright protection
2. Source tracking
3. Broadcast monitoring
5. Piracy deterrence etc.,
The categorization of watermarking scheme into four phases is based on the kind of information to be watermarked; this information can be any of (Sinha et al., 2017):
1. Video watermarking
2. Image watermarking
3. Text watermarking
4. Audio watermarking
Image watermarking was considered in this research work but any of the other watermarking types can be used.
A model of bits fixed into a digital audio, video, image, or text file that is unambiguously used to identify the owner of a specific image is refered to as watermark and its main steps are embedding and extraction (Cherian & Mereena, 2016; Nyeem et al., 2014).
1. Embedding Processing: Embedding is a method of inserting a watermark within the host image in order to create the watermarked image and the process is carried out at the sender‘s side (Navas et al., 2008)
A typical watermark embedding is represented as follows (Chandramouli et al., 2002):
( ) ( )
X is the original image
W is the watermark information being embedded
k is the user‘s insertion key
E represents the watermark insertion function
X‘ depicts the watermark variant
2. Extracting Processing: Extracting is the method of recovering the watermark and the host image from the watermarked image and this process happens at the receiver‘s end
A generic watermark extraction is represented as follows (Chandramouli et al., 2002):
̂ ( ̂) (1.2)
̂ represents the possible corrupted watermarked image
denotes the extraction key
D depicts the watermark extraction /detection function
̂ represents the extracted watermark information
The two processes are shown in Figures 1.1 and 1.2 respectively.
Figure 1.1: Watermark Embedding Process (Chanda & Choudhury, 2016)
Figure 1.1 depicts the watermarking embedding process where the original image represents the carrier signal or message that hosts the watermark while the watermark is the hidden data. The watermarked image is the media which contains the watermark and is generated by using an embedding function to insert the watermark into the original image. This operation is executed from the sender‘s end (Chanda & Choudhury, 2016). ORIGINAL IMAGE WATERMARK WATERMARK EMBEDDING PROCESSING WATERMARKED IMAGE
Figure 1.2 depicts the watermarking extraction phase where the watermarked image undergoes extraction process with a view to recovering the original watermark at the receiver‘s side.
Figure 1.2: Watermark Extraction Process (Chahal & Khurana, 2013)
Figure 1.3 depicts the complete representation of the components of a watermarking process which comprises of embedding, detection and extraction processes.
Figure 1.3: Watermarking System (Tao & Eskicioglu, 2004)
Detection is the procedure deployed for identifying whether the given media contains the watermark or not (Tao & Eskicioglu, 2014). Attack is an artificial process used intentionally or non-intentionally to duplicate or modify the watermark within the cover image. This modifies the watermarked image and destroys or alters watermark in the data. Key is a bit of information or parameter that controls the embedding and extraction processes (Tao & Eskicioglu, 2014). WATERMARKED IMAGE WATERMARK EXTRACTION PROCESS ORIGINAL IMAGE WATERMARK
where: I depicts the host image, W Denotes the watermark image, IW represents the watermarked image, IW* represents the distorted image, K depicts embedding and extraction key
The two key groups of digital image watermarking are as follows (Sathik & Sujatha, 2012):
1. Visible digital image watermarking technique
2. Invisible digital image watermarking technique
Information is perceptible in the multimedia content which identifies the owner of the document in detectible watermarking. The visible watermarking should be perceptible (easy to discern the concealed data) (Santhi & Arulmozhivarman, 2013).
In invisible watermarking information is inserted as digital data to multimedia content, such that, it is imperceptible to observers. The invisible watermarking requirements are imperceptibility and robustness. The invisible watermarks are categorized as follows (Chandramouli et al., 2002):
1. Fragile watermarks
2. Semi fragile watermarks
3. Robust watermarks
Several techniques have been proposed for copyright protection of digital images and these are categorized into two domains as follows (Sathik & Sujatha, 2012):
1. Spatial domain techniques
2. Transform domain techniques
In spatial domain techniques the pixel values of the host image are changed directly by inserting the watermark bits (Araghi et al., 2016). The spatial domain techniques are simply implemented with little cost of operation, quick, computationally uncomplicated and
straightforward while there is no need for cover image to be transformed. Nevertheless, they are vulnerable to simple image processing operations such as noise, compression and filtering due to watermark insertion in selected locations of the image or other geometric attacks. The spatial domain techniques comprise of the following (Lai & Tsai, 2010; Zheng, Liu, Zhao, & Saddik, 2007):
1. Least significant bit (LSB)
2. Spread spectrum technique (SSM)
3. Correlation based technique (CBT)
In transform domain technique the coefficients of the host image is modified where the watermark is inserted for embedding in transform domain. At first the cover image is transformed and then the watermark is inserted to the coefficients of the transformed image. In order to retrieve the original image, an inverse transform of the modified coefficients needs to be performed (Araghi et al., 2016). The transform domain techniques comprise of the following (Rahman, 2013 and Cherian & Mereena, 2016):
1. Discrete Orthonormal Stockwell Transform (DOST)
2. Singular Value Decomposition (SVD)
3. Lifting Wavelet Transform (LWT)
4. Discrete Stockwell Transform (DST)
5. Discrete Cosine Transform (DCT)
6. Discrete Wavelet Transform (DWT)
7. Discrete Fourier Transform (DFT)
8. Fast Fourier Transform (FFT)
The information concealing and recovery operations of these transform algorithms are comparatively complicated, however, due to their strong anti-attack abilities, they are suitable for the copyright protection of multimedia data (Araghi et al., 2016). These methods in handling image and common signal processing attacks achieve higher imperceptibility and robustness, although, the cost of computation is higher than that of the spatial domain watermarking schemes (Cherian & Mereena, 2016).
Watermarking system is classified based on three vital criteria, namely the type of domain, watermark and information required in the recovery or extraction process. The classification according to these criteria is shown in Table 1.1
Table 1.1: Classification of Digital Watermarking Technique (Araghi et al., 2016):
Pixels values are adjusted to enclose the secret message.
Transform constants are modified to insert the classified message. Present popular transforms are DCT, DWT, and DFT etc.
Pseudo random number (PRN) sequence (having a normal distribution with zero mean and unity variance)
Let the detector to statistically determine the existence or nonexistence of a watermark. A PRN series is created by serving the generator with an underground seed.
The watermark is essentially reassembled, and its visual value is evaluated.
Both the original image and the undisclosed key(s) are require for watermark extraction process
The watermark and the secret key(s) are require for the extraction process
Only the secret key(s) is require for the extraction process
The use of digital data (image, audio, video and text) is rapidly rising with the high growth and widespread use of the Internet and information technology. Due to this fact, tampering
and illegal distribution of digital contents is inevitable and as such it becomes imperative to devise mechanisms to protect such contents. It has been established that present copyright laws are insufficient in dealing with the security of digital data (Chandramouli et al., 2002). Information hiding and copyright protection have also become vital challenges on the issue of sharing digital data over public network; this is the principal motivation for this research work. In order to address these challenges watermarking technology is adopted. Several researchers had worked in the area of watermarking for its usefulness (Al-Mansoori & Kunhu 2012, Gattani & Warnekar, 2014; Chanda & Choudhury 2016; Ouazzane et al., 2017). Work in this area has led to development of several watermarking techniques such as spatial domain based techniques and transform domain based techniques. Image watermarking scheme should possess the requirements of imperceptibility between the watermarked image and original image and robustness (Tao et al., 2014). Robustness is the ability of the digital watermarking method to withstand common image manipulations like compression, filtering, rotation, scaling, cropping and other digital signal processing operations (Tao et al., 2014, Mehto & Mehra, 2015). Normally used frequency domain transforms include the DFT, DWT, DCT, and SVD (Chan & Cheng, 2004). Currently DOST has been developed and applied to image texture characterization, image compression, decomposition. DWT has the capability of producing temporal resolution and captures frequency and location information, these properties of DWT make its suitable for the application of de-noising, compression and phase analysis. Although, the self-similarity restraint amongst the wavelet basis functions ruins the phase information; as such coefficients supply single locally referenced phase information. SVD is a transform technique that addressed the challenge associated with DWT transform technique. SVD offers a robust method of watermarking with little or no distortion based on
the singular value of the image. This makes it resilient to noise and geometric attack, even though, it suffers from false positive outcome. DOST has the ability to retain phase reference information of the signal absolutely, that is, the reference is constantly at zero time. With this, it gives the accumulation of the coefficients for a static frequency that yields the precise Fourier coefficient for that frequency. However, the non-redundancy nature ascribable to DOST, make it provides a preferably coarse time–frequency representation with its frequency resolution proportionally scaled to the logarithm of the frequency. Such an illustration may not be constantly simple to interpret and be sufficient to disclose all the details within a specific signal. Therefore, the cascade of DOST, DWT and SVD is utilized in this work in order to compensate for the deficiencies of each other, with a view to achieving an effective watermarking technique. The watermark is encrypted with a private key to further enhance the security of the watermarking algorithm. Therefore, the proposed watermarking technique is expected to achieve an optimal balance between imperceptibility and robustness when subjected to attacks.
1.3 Significance of Research
In recent years the nature of digital network systems means that digital contents can be duplicated and shared with ease to large spectrum of people with no cost over internet. Due to current advances in information technologies, these multimedia data can easily be downloaded, duplicated and manipulated by hackers in order to modify the original contents (Kalarikkal et al., 2017). The major challenges is how to protect the ownership right of these multimedia contents transmitted over internet and prevent duplication of the contents, which still remains a major challenge to researchers. In addressing these challenges, researchers introduced a copy deterrence mechanism called digital watermarking. In order for digital
watermarking technique to be useful, it must satisfy imperceptibility and robustness requirements for protection. Therefore, this thesis is focused on the development of a robust digital image watermarking algorithm in DOST domain by employing DWT in conjunction with SVD based on photon polarization with a view to protecting the copyright of multimedia data, with the hope of striking a balance between imperceptibility and robustness.
The unique features of the developed technique are as follows:
1. The watermarking algorithm is blind since the original cover image is not involve during the recovery process.
2. The watermarking algorithm is robust against variety of attacks.
3. The watermark is invisible and causes no distortion to the cover image.
1.4 Statement of Problem
The rising pace in the transmission of multimedia contents (such as image, audio, video and text) over secured and unsecured communication media poses a lot of security and privacy concerns to both the sender and the receiver. It has been established that present copyright laws are insufficient for dealing with the security of digital data (Chandramouli et al., 2002). This has led to the development of a copy deterrence and protection mechanism called digital watermarking (Kavitha & Shan, 2016). Most works reported in literature using different techniques only concentrated on achieving the watermark robustness by compromising image fidelity or vice versa. In other words, those techniques were based on some sort of trade-off between robustness and imperceptibility of the watermarked image. Hence, this research sets to develop a novel approach of robust digital image watermarking technique in DOST domain using DWT and SVD. The scaling factor, which is critical to imperceptibility and robustness, is optimally obtained using Particle Swarm Optimization (PSO). The secret key is obtained
using quantum key distribution (QKD) based on the principle of photon polarization in order to mitigate the effect of both signal processing manipulation and geometric attacks, with the hope of simultaneously satisfying both robustness and imperceptibility requirements.
1.5 Aim and Objectives
This research is aimed at developing a robust digital image watermarking technique in DOST domain based on the principle of photon polarization in order to strike a balance between imperceptibility and robustness to obtain the copyright protection of a digital image.
In order to accomplish the stated aim, the following objectives were adopted:
1. To develop a digital image watermarking algorithm in 2D DOST domain using DWT and SVD for decomposition of the cover image and watermark for both embedding and extraction processes.
2. To develop a scheme for the generation of the secret key from quantum key distribution using the principle of photon polarization based on the BB84 protocol
3. To determine the optimal scaling factor by applying particle swarm optimization (PSO)-based approach required to ascertain the strength of the embedded watermarked image
4. To generate a watermarked image by performing inverse DWT and 2D DOST on the transformed coefficients (obtained from 1 modified by the key and scaling factor obtained from 2 and 3 respectively). Extraction of the watermark is to be carried out by performing the inverse of the process described.
5. To validate the developed technique by comparing with the ITU benchmark and techniques developed by Bajracharya & Koju (2017 in terms of robustness and
imperceptibility using peak signal to noise ratio (PSNR), normalized-correlation coefficient (NCC) and structural similarity index (SSIM) as performance metrics.
1.6 Scope of the Research
The scope of this research work which developed a new data hiding algorithm (digital watermarking technique) inspired by the proliferation of duplicating and modifying of intellectual properties on daily basis without the consent of the rightful owner. The scope that encapsulates this developed algorithm is enumerated as follows:
1. The research only considered DOST, DWT and SVD for images transformation and as such no other transform domain techniques were considered.
2. The research only considered affine transform and XORed operation technique in encryption and decryption of the watermark image.
3. An optimal scaling factor selected by the application of PSO only for embedding the watermark is considered in this thesis.
4. The research only considered colour images
1.7 Thesis Organization
This thesis consists of five chapters, and is organized as follows: The introductory part of the thesis is discussed in chapter one. Chapter two provides an overview of the pertinent fundamental concepts in addition to a review of similar works. Chapter three describes the detailed steps of the methodology while results and discussions are presented in chapter four. Summary, conclusion and recommendations make up chapter five. Appendices and relevant references are presented at the end of this thesis.
GET THE COMPLETE PROJECT»
Do you need help? Talk to us right now: (+234) 08060082010, 08107932631, 08157509410 (Call/WhatsApp). Email: firstname.lastname@example.org