Protecting Digital Content 1
1. INTRODUCTION
1.1 Brief Introduction
A watermark is a recognizable image or pattern in paper that appears as various shades
of lightness/darkness when viewed by transmitted light (or when viewed by reflected
light, a dark background), caused by thickness or density variations in the paper. There
are two main ways of producing watermarks in paper; the dandy roll process and
complex cylinder mold process. Watermarks are often used as security features of bank
notes, passports, postage stamps and other documents to prevent counterfeiting.
Watermark is very useful in the examination paper because it can be used for dating,
identifying sizes, mill trademarks and locations, and the quality of a paper. We are
living in the era of information where billions of bits of data is created in every fraction
of a second and with the advent of internet, creation and delivery of digital data (images,
video and audio files, digital repositories and libraries, web publishing) has grown
many fold. Since copying a digital data is very easy and fast too so, issues like,
protection of rights of the content and proving ownership, arises. Digital watermarking
came as a technique and a tool to overcome shortcomings of current copyright laws for
digital data. The specialty of watermark is that it remains intact to the cover work even
if it is copied. So to prove ownership or copyrights of data watermark is extracted and
tested. It is very difficult for counterfeiters to remove or alter watermark. As such the
real owner can always have his data safe and secure. Our aim is to study different
watermarking techniques and all types of attack. Counterfeiters try to degrade the
quality of watermarked image by attacking an image (generally attacks are scaling,
compression and rotation of watermarked image). By attacking watermarked image it
becomes very difficult to recover watermark back from the watermarked image and
even if it extracted one may no longer use it to prove the ownership and copyrights. So
our main idea is to find such regions, also known as patches, in an image which are
very stable and resistant to attacks. The term watermark may have been derived from
the German term, sermarke. The term is actually a misnomer, in that water is not
especially important in the creation of the mark. It was probably given because the
marks resemble the effects of water on paper. At the beginning of 1990 the idea of
digital watermarking has emerged, it embedding imperceptible information into audio
visual data. The first watermarking methods were proposed for digital images by
Caronni in 1993. Digital watermarks have mainly three application fields: data
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 2
monitoring, copyright protection and data authentication. The art of watermarking was
invented in China over one thousand years earlier. The marks were made by adding thin
wire patterns to the paper moulds. The paper would be slightly thinner where the wire
was thicker and hence more transparent. The meaning and purpose of the earliest
watermarks are uncertain. They may have been used for practical functions such as
identifying the moulds on which sheets of papers were made, or as trademarks to
identify the paper maker.
1.2 PROBLEM DEFINITION
Digital image watermarking is a technique used to embed hidden information
(watermark) into digital images to ensure authenticity, copyright protection, and
integrity verification. The project aims to explore and implement watermarking
techniques using a combination of Discrete Wavelet Transform (DWT), Discrete
Cosine Transform (DCT), and Binary Firefly Optimization (BFO) for optimization
1.3 OBJECTIVE AND SCOPE
Objectives
1. Implementation of Techniques
Implement digital image watermarking using DWT.Implement digital image
watermarking using DWT and DCT.Implement digital image watermarking using
DWT, DCT, and optimize parameters using BFO.
2. Comparison
Compare the effectiveness of watermarking using DWT alone, DWT combined with
DCT, and DWT combined with DCT optimized by BFO. Evaluate the robustness of
each technique against common attacks (e.g., noise addition, compression).
3. Performance Evaluation
Measure the imperceptibility of the watermark (visual quality of the watermarked
image). Assess the robustness of the watermark against various attacks to determine the
reliability of each method.
4. Optimization
Apply BFO to optimize the embedding parameters in DWT and DCT to enhance
watermark robustness and imperceptibility.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 3
Scope The watermarking technique cannbe further extended by implementing at
Swarm optimization method after the DWT.
1.4 BACKGROUND AND ALTERNATIVES
Digital image watermarking is a technique employed to embed imperceptible
information, known as a watermark, into digital media for various purposes such as
copyright protection, authentication, and content integrity verification. The challenge
lies in achieving a balance between robustness against attacks and maintaining visual
quality (imperceptibility) of the watermarked image.
Techniques Involved
Discrete Wavelet Transform (DWT)
DWT decomposes an image into different frequency bands, allowing for selective
embedding of the watermark in specific frequency domains. This technique is known
for its ability to achieve a good compromise between imperceptibility and robustness.
Discrete Cosine Transform (DCT)
DCT is another widely used transform technique that converts spatial image data into
frequency components. It is particularly effective in compressing and embedding
watermarks in the frequency domain, although it may be more susceptible to certain
types of attacks compared to DWT.
Binary Firefly Optimization (BFO)
BFO is a metaheuristic optimization algorithm inspired by the flashing behavior of
fireflies. It has been applied to optimize various parameters in digital watermarking,
such as embedding strengths, scaling factors, and locations for embedding, aiming to
enhance watermark robustness and imperceptibility.
Project Focus
This project focuses on exploring the synergies between DWT and DCT for digital
image watermarking, enhanced through optimization using BFO. By combining these
techniques, the project aims to achieve improved robustness against common attacks
while minimizing perceptual distortion in the watermarked image.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 4
Alternatives Considered
Frequency Domain Techniques
Other frequency domain techniques such as Fourier Transform, and its variants could
be considered for comparison to evaluate their suitability in watermarking applications.
Spatial Domain Techniques
Techniques that embed watermarks directly in the spatial domain, such as LSB (Least
Significant Bit) embedding, could serve as benchmarks to compare against frequency
domain approaches.
Optimization Algorithms
Apart from BFO, alternative optimization algorithms like Particle Swarm Optimization
(PSO), Genetic Algorithms (GA), or Simulated Annealing could be explored for
optimizing watermark embedding parameters.
Hybrid Approaches
Hybrid approaches combining multiple transforms (e.g., DWT-DCT hybrid) or
optimization techniques (e.g., combining BFO with PSO) could potentially offer further
enhancements in watermark robustness and imperceptibility.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 5
2. LITERATURE SURVEY
Digital watermarking plays a crucial role in protecting digital images from unauthorized
copying and tampering. Both Discrete Wavelet Transform (DWT) and Discrete Cosine
Transform (DCT) offer effective avenues for embedding watermarks in the frequency
domain, each with their own advantages and limitations. Here's a look at the literature
on these techniques and their combined approach (DWT+DCT).
DWT-based Watermarking
DWT decomposes the image into sub-bands capturing spatial and frequency
information. This allows for embedding watermarks in specific sub-bands based on
their characteristics.
Papers like "[Combined DWT-DCT digital image watermarking]" by Nosrati et al.
(2007) demonstrate watermark embedding in the DWT domain, achieving good
imperceptibility and robustness against compression attacks.
DWT's strength lies in its ability to localize information, making it suitable for
watermarking with good tamper detection capabilities.
DCT-based Watermarking
DCT transforms an image into the frequency domain, where high-frequency
components hold the detail suitable for watermark embedding.
Research by Cox et al. (1996) in "[A secure, robust watermark for multimedia]"
explores DCT-based watermarking, achieving robustness against common attacks like
noise addition.
DCT excels in energy compaction, concentrating information in specific coefficients,
making it suitable for robust watermarking.
DWT+DCT Watermarking (Hybrid Approach)
Combining DWT and DCT offers the potential to leverage the strengths of both
techniques. Papers like "[Analysis On Digital Image Watermarking Using Dct-Dwt
Techniques]" by Pranay et al. (2018) showcase DWT+DCT watermarking, achieving
improved robustness compared to using DWT alone.
This approach allows for embedding watermarks in both spatially localized (DWT) and
frequency-domain (DCT) components, potentially enhancing imperceptibility and
robustness.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 6
Benefits of DWT+DCT
Improved robustness against a wider range of attacks by utilizing the strengths of both
transforms.Potential for increased watermark capacity by embedding in both DWT and
DCT domains.Possibility of achieving better imperceptibility by carefully selecting
embedding locations. However, the combined approach can also lead to increased
computational complexity compared to using DWT or DCT alone.
2.1 Problem Statemen
Proposed System
In our proposed work the message is hide in the cover image. Objective is to achieve invisible
watermarking such that after embedding the message, message shouldn’t be observable. The
method adopted for this purpose is the combination of three methods which are used in
watermarking individually. But every method has its own limitations so in our work these are
combined to overcome their limitations. The method adopted here is Discrete Wavelet
transform (DWT) along with Discrete cosine transform (DCT), optimized with bacterial
foraging optimization (BFO). The performance criteria’s for image watermarking are PSNR
(Peak signal to noise ratio) value, NCC (normalized cross correlation) and IF (image fidelity).
The value of PSNR and NCC must be high for good embedding of message. The embedding
of message by any selected method comes with a constraint that the message should be
recovered at the receiver end clearly and in that case validation can be done by the NCC as
normalized cross correlation between the original message and recovered message must be
high. So, a gain factor in the proposed embedding process is introduced, discussed in next
section, which decides the depth of message hiding and retrieval also. But it is also required
that embedding should be invisible and robust also to any type of attack or noise introduced
during transmission of image. High PSNR value guarantees robustness of watermarked image.
So gain factor value must be selected so that a balance between the PSNR and NCC can be
managed. To set the optimum value of gain factor bacterial foraging optimization is used in our
work. The BFO as discussed in previous chapter minimize the objective function value to get
the best location of E.Coli bacteria. So the task is to formulate the objective function to achieve
the best gain value. For this purpose inverse of normalized cross correlation has been considered
the parameter which is to be minimized. Initially random gain value is selected and that is
passed to the embedding and retrieval process of message using DWT and DCT watermarking
process. At the watermarked image by this process various noises like Gaussian noise, salt &
pepper noise, speckle noise and Poisson noise have been added and message is recovered from
these noisy images. NCC between the original message and recovered from these noisy
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 7
watermarked images have been found out. Then the inverse of sum of all these NCCs is
considered as the objective function of BFO
2.2 Problem Identification
Challenges
Imperceptibility vs. Robustness Trade-off Achieving a balance between embedding
a watermark that is imperceptible to human perception and robust enough to withstand
various attacks (e.g., noise addition, compression) remains a significant
challenge.DWT offers multi-resolution analysis for selective embedding in different
frequency subbands, but maintaining imperceptibility in high-frequency components
can be challenging.DCT-based methods embed watermarks efficiently in low-
frequency coefficients but may lack robustness under certain types of attacks.
Security and Authentication Requirements Ensuring the security and authentication
of embedded watermarks against malicious tampering or removal is crucial.Systems
must be resilient to attacks that attempt to alter or remove watermarks without
degrading the quality or integrity of the original image.
Optimization of Embedding Parameters Optimization techniques such as Genetic
Algorithms (GA), Particle Swarm Optimization (PSO), or Binary Firefly Optimization
(BFO) play a crucial role in enhancing watermarking systems.Balancing embedding
strength, scaling factors, and localization of watermarks within DWT and DCT
coefficients requires efficient optimization strategies to improve robustness and
imperceptibility simultaneously.
Opportunities
Integration of DWT and DCT Hybrid approaches combining DWT for multi-
resolution analysis and DCT for efficient frequency domain representation offer
opportunities to enhance watermarking performance.Leveraging the complementary
strengths of both transforms can lead to systems that achieve superior robustness and
imperceptibility compared to individual techniques.
Advancements in Optimization Techniques Continued advancements in
optimization algorithms like BFO present opportunities to fine-tune embedding
parameters dynamically based on image content and security requirements.
Optimization algorithms can improve the efficiency and effectiveness of watermarking
systems, making them more adaptable to diverse application scenarios.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 8
Emerging Applications and Use Cases Digital image watermarking is expanding
beyond traditional copyright protection to include applications in multimedia forensics,
content authentication, and secure data embedding in IoT (Internet of Things)
environments.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 9
3. METHODOLOGY
Implementation Phase
Implementation of watermark embedding and extraction using DWT.
Extension to incorporate DCT for frequency domain embedding.
Integration of BFO to optimize watermark parameters like scaling factors and
embedding strengths.
Evaluation Phase
Performance evaluation metrics: Peak Signal-to-Noise Ratio (PSNR), Structural
Similarity Index (SSIM).
Testing under various conditions: noise addition, compression, resizing, and geometric
transformations.
Comparison Phase
Comparative study of DWT, DWT+DCT, and DWT+DCT optimized by BFO.
Analysis of experimental results to highlight strengths and weaknesses of each
technique.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 10
4. IMPLEMENTATION
1. Environment Setup Set up the development environment with necessary tools
and libraries: MATLAB for DWT implementation.Python with NumPy for DCT
implementation. Integrated development environment (IDE) such as PyCharm or
Jupyter Notebook for coding and experimentation.
2. Implementation of DWT-Based WatermarkingImage Preprocessing
Load digital images from the dataset. Convert images to grayscale or RGB format as
required.Normalize pixel values to a standard range (e.g., 0-255).
Discrete Wavelet Transform (DWT) Select a DWT implementation method (e.g.,
built-in MATLAB functions like dwt2).Choose a suitable wavelet basis (e.g., Haar,
Daubechies) and decomposition levels.Decompose images into approximation (LL),
horizontal detail (HL), vertical detail (LH), and diagonal detail (HH) coefficients.
Watermark Embedding Embed the watermark into selected DWT coefficients (e.g.,
HH or LL subbands) using embedding strength and scaling factors.Adjust embedding
parameters to achieve a balance between imperceptibility and robustness against
attacks.
Watermark Extraction Extract the watermark from watermarked images using the
reverse process of DWT.Verify the integrity and correctness of extracted watermarks
using validation techniques.
1. Implementation of DCT-Based Watermarking:
Image Preprocessing Load digital images and convert them to grayscale or RGB
format. Normalize pixel values to facilitate DCT processing.
Discrete Cosine Transform (DCT) Implement DCT using Python with NumPy for
efficient frequency domain representation. Perform DCT on image blocks or the entire
image to obtain DCT coefficients.
Watermark Embedding Embed the watermark into low-frequency DCT coefficients
to ensure robustness against compression and maintain perceptual quality. Adjust
embedding parameters (e.g., quantization steps) to optimize embedding capacity and
detectability.
Watermark Extraction Extract the watermark from watermarked images by reversing
the embedding process. Validate extracted watermarks using correlation or error
checking methods.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 11
4. Integration of DWT+DCT Combined Approach Develop a hybrid
watermarking technique that integrates DWT and DCT methodologies: Use DWT for
multi-resolution analysis and selective embedding in frequency subbands. Employ
DCT for efficient representation of image data and embedding in low-frequency
coefficients.Optimize embedding parameters using techniques such as Genetic
Algorithms or Particle Swarm Optimization to enhance watermark robustness and
imperceptibility.
5. Optimization Techniques Apply Binary Firefly Optimization (BFO) or other
optimization algorithms to refine embedding parameters dynamically: Adjust scaling
factors, embedding strength, and localization of watermarks within DWT and DCT
coefficients. Iterate optimization processes to maximize watermark robustness against
common attacks.
6. Performance Evaluation Conduct comprehensive performance evaluation
using objective metrics: Calculate Peak Signal-to-Noise Ratio (PSNR) and Structural
Similarity Index (SSIM) to assess image quality and fidelity. Measure embedding
capacity and detectability using Bit Error Rate (BER) under various attack scenarios
(e.g., JPEG compression, noise addition).
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 12
5. ADVANTAGES AND DISADVANTAGES
Advantages
Multi-Resolution Analysis DWT provides multi-resolution decomposition, allowing
selective embedding of watermarks in different frequency subbands. This enhances
robustness against various attacks while maintaining perceptual quality.
Localization DWT enables precise localization of watermarks within image
components, minimizing visual degradation and preserving image fidelity.
Flexibility Adaptability to different image sizes and types, making it suitable for
diverse applications in multimedia security and content authentication.
Efficient Representation DCT efficiently represents image data in the frequency
domain, particularly suitable for embedding watermarks in low-frequency coefficients.
Compression Robustness Watermarks embedded in DCT coefficients exhibit
robustness against compression algorithms like JPEG, ensuring integrity and quality
preservation.
Computational Efficiency Fast computation of DCT makes it practical for real-time
applications and large-scale watermarking tasks.
Complementary Strengths Integration of DWT and DCT leverages their
complementary characteristic multi-resolution analysis of DWT enhances localization
and robustness, while DCT ensures efficient frequency domain representation and
compression resilience.
Improved Robustness Enhanced resistance against a broader range of attacks (e.g.,
noise addition, geometric transformations) compared to individual transforms alone.
Optimization Potential Opportunities for optimizing embedding parameters using
hybrid techniques and optimization algorithms, leading to improved watermarking
performance.
Disadvantages
Discrete Wavelet Transform (DWT) Perceptual Quality: Embedding watermarks in
high-frequency DWT coefficients can degrade perceptual quality, making the
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 13
watermark more noticeable to human observers. Complexity: Selection of wavelet basis
and decomposition levels requires expertise and experimentation, impacting
implementation complexity and computational resources. Vulnerability to Certain
Attacks: Specific attacks targeting DWT-based watermarks, such as geometric
transformations in certain subbands, may compromise robustness.
Limited Frequency Localization DCT lacks the frequency localization capabilities
of DWT, potentially reducing the precision of watermark embedding and detection in
specific frequency bands.
Sensitivity to Scaling Watermark robustness in DCT can be sensitive to scaling
operations and quantization steps, affecting detection reliability under certain
conditions.
Trade-offs in Imperceptibility Balancing imperceptibility with robustness in DCT-
based watermarking may require trade-offs, impacting visual quality in highly
compressed images.
Complex Integration Implementing and optimizing combined DWT+DCT
techniques requires careful integration of algorithms and parameters, increasing
development complexity.
Parameter Sensitivity Sensitivity of hybrid techniques to parameter settings (e.g.,
scaling factors, embedding strengths) may necessitate iterative optimization processes,
potentially increasing computational overhead.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 14
6. TECHNOLOGY
MATLAB:
MATLAB is short for “matrix laboratory.” It is a high-level programming language and
platform that works with matrices and arrays rather than individual numbers. Designed
for scientists and engineers, MATLAB is often used in textbooks as an instructional tool
for college-level mathematics, science, and engineering. It allows for the most direct and
natural expression of matrix and array mathematics.
Utilized for implementing Discrete Wavelet Transform (DWT) algorithms. Offers
built-in functions (dwt2, idwt2) for efficient wavelet decomposition and reconstruction.
Provides a user-friendly environment for prototyping and testing watermarking
techniques.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 15
7. CONCLUSION AND FUTURE SCOPE
The effectiveness of the whole scheme is proven through simulation results like -
1) PSNR quality assessment objectives are achieved
2) watermarked image have very good visual quality
3) no auxiliary data is required for quality estimation (only embedded watermarks and
test images are needed).
In this work, a still image watermarking scheme with high robustness in the frequency
domain is applied. The proposed scheme tests only image rather than audio or video.
This algorithm can be used for data hiding in many applications such as authentication
and copyright protection. In this paper, a general coding-type framework which
provides useful and constructive tools in the analysis and design of watermarking
system is used. That particularly demonstrates the effectiveness of watermarking
approach in achieving design objectives such as robustness, capacity, security, and
implementation efficiency.
FUTURE SCOPE
The watermarking technique that is given in this paper can be further extended by implementing
at Swarm optimization method after the DWT.
In future, work may be extended on different media like video, audio etc by using this approach.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
Protecting Digital Content 16
8. REFERENCES
1) Abdelaziz I. Hammouri, Basem Alrifai and Heba AlHiary,” An Intelligent Watermarking
Approach Based Particle Swarm Optimization in Discrete Wavelet Domain” IJCSI
International Journal of Computer Science Issues, Vol. 10, Issue 2, No 1, March 2013
2) Ray-Shine Run, Shi-Jinn Horng, Jui-Lin Lai, TzongWang Kao, Rong-Jian Chen,” An improved
SVD-based watermarking technique for copyright protection” Expert Systems with
Applications 39 (2012)
3) Sonil Sood, Ajay Goyal,” Watermarking Relational Databases using Genetic Algorithm &
Bacterial Foraging Algorithm” International Journal of Information & Computation
Technology, Volume 4, Number 17 (2014)
4) Hsiang-Cheh Huang, Yueh-Hong Chen, Ajith Abraham,” Optimized Watermarking Using
SwarmBased Bacterial Foraging” Journal of Information Hiding and Multimedia Signal
Processing, Volume 1, Number 1, January 2010.
5) Sonil Sood,” Digital Watermarking Using Hybridization of Optimization Techniques:A
Review” International Journal of Computer Science and Information Technologies, Vol. 5
(4) , 2014.
6) P. Surekha and S. Sumathi,” Implementation Of Genetic Algorithm For A Dwt Based Image
Watermarking Scheme” Ictact Journal On Soft Computing: Special Issue On Fuzzy In
Industrial And Process Automation, July 2011, Volume: 02, Issue: 01.
7) Zhicheng Wei, Hao Li, Jufeng Dai ,Sashuang Wang,” Image Watermarking Based On
Genetic Algorithm” IEEE 2006
8) Mahmoud Elnajjar, A.A Zaidan, B.B Zaidan, Mohamed Elhadi M.Sharif and
amdan.O.Alanazi,” Optimization Digital Image Watermarking Technique for Patent
Protection” Journal Of Computing, Volume 2, Issue 2, February 2010.
V.P. Dr P.G.H.C.E.T Department of Computer Application 2023-2024
Vijayapur (MCA)
