Conducted Under IEEE MTT-S & IEEE-AESS Chapter SBC11474
Joint Delhi Chapter IEEE PES-IAS Delhi Section · PE31/IA34
ABV-Indian Institute of Information Technology and Management, Gwalior
Free Webinar · Image Processing · Digital Security

Color Image Watermarking with
Watermark Authentication Against False Positive Detection Using SVD

Join Dr. Neha Singh for an expert-led session on securing digital images using SVD-based watermarking. Explore robust embedding techniques, cover-dependent key authentication, and practical defenses against ambiguity attacks — grounded in published peer-reviewed research.

May 10, 2026
12:00 p.m. Pacific Time
Live Online • Free Entry
00Days
00Hours
00Min
00Sec
Reserve Your Free Seat

Featured Speaker

Published researcher, Senior IEEE Member, and 21-year educator in Electronics & Communication Engineering

Dr. Neha Singh
LinkedIn Google Scholar Scopus ResearchGate
Dr. Neha Singh
Ph.D. · Manipal University Jaipur (2020) · VLSI Design & Nanodevices
Department of Electronics & Communication Engineering · Senior Member, IEEE

Dr. Neha Singh is a dynamic teaching and research professional with 21 years of experience educating undergraduate and postgraduate engineering students at Manipal University Jaipur, Rajasthan, India.

Her research spans Image Processing, Machine Learning, VLSI Design, and Nanodevices. She has published numerous indexed papers, co-authored engineering textbooks, edited three volumes with international publishers, and is currently supervising Ph.D. scholars. She is a Senior Member of IEEE.

Image Processing
Digital Security
VLSI Design
Machine Learning
21+Years Teaching
49+Publications
IEEESenior Member
3Edited Books

Research Paper in Focus

This webinar is built directly on Dr. Singh's peer-reviewed publication. Here is a deep look at the core technique presented.

Published Research
Cover Dependent Watermarking Against Ambiguity Attacks
Procedia Computer Science · Third International Conference on Computing and Network Communications (CoCoNet'19) · Elsevier, CC BY-NC-ND 4.0
The proposed work embeds Principal Components of two watermarks into horizontal and vertical sub-band coefficients of a 3-level DWT of the cover image. Cover-and-watermark-based keys are generated for verification during extraction — ensuring that knowing the keys alone is insufficient to extract the watermarks without the correct DWT coefficient combination.
Digital Image Watermarking Singular Value Decomposition DWT False Positive Detection Ambiguity Attacks HSV Colour Model Principal Components

Core Technique — Embedding Steps

1
Convert the 512×512 colour cover image to HSV colour model; extract the Value plane for processing.
2
Compute Principal Components (PC = UW × SW) of two 64×64 grayscale watermarks via SVD.
3
Apply 3-level DWT on the Value plane; obtain H3 and V3 sub-band coefficients.
4
Perform SVD on H3 and V3; update singular values with scaled PCs using a strength factor (sf).
5
Re-apply SVD on the updated singular value matrices to generate 4 cover-dependent keys (Key1–Key4) stored for extraction.
6
Reconstruct updated H3nu and V3nu, apply inverse DWT, and rebuild the watermarked image in HSV → RGB.

Performance Results

43.98
PSNR (dB) — no attack
1.00
NC — perfect extraction
0.96+
NC after median filter
0.99+
NC after resize (2→0.5)

Why This Beats Standard SVD Watermarking

  • ✓ Embeds PCs — not just SVs — so fake watermark extraction is structurally impossible
  • ✓ Keys are cover-dependent: processed covers generate wrong keys and fail authentication
  • ✓ Dual watermarks in H3 and V3 bands — at least one survives most attack scenarios
  • ✓ Outperforms Ansari & Pant [2017] and Shivani & Senapati [2018] on most attacks
  • ✓ Independent sf(H) and sf(V) allow fine-tuning of the PSNR–robustness trade-off
Authors
Neha Singh · Sandeep Joshi · Shilpi Birla
Department of Electronics & Communication Engineering and Department of Computer Science & Engineering, Manipal University Jaipur, Rajasthan-303007, India

Frequently Asked Questions

Everything you need to know before joining the webinar.

What is the false positive problem in SVD watermarking?
In standard SVD-based watermarking, only the Singular Values (SVs) of the watermark are embedded into the cover image. During extraction, these SVs must be combined with the watermark's singular vectors for reconstruction. An attacker can substitute the singular vectors of any fake watermark and produce a plausible extracted result — falsely claiming ownership. This webinar presents a technique that eliminates this flaw by embedding Principal Components instead.
What is "cover-dependent" watermarking and why does it matter?
The four keys used to embed and extract watermarks are generated from a combination of the original cover image and the watermarks. If an attacker attempts to claim ownership using a processed version of the cover (histogram equalized, filtered, etc.), the keys generated from that fake cover will not match, and watermark extraction will fail — proving the claim is invalid.
Do I need prior knowledge of SVD or DWT to follow the webinar?
Basic familiarity with image processing (pixel structures, colour models) is helpful, but not mandatory. The session starts from the foundations of Digital Image Watermarking and builds up to the proposed technique step by step — suitable for advanced students, researchers, and working engineers alike.
What makes this technique better than existing methods?
The technique was compared against Ansari & Pant (2017) and Shivani & Senapati (2018) at equivalent PSNR values (~38 dB and ~42.6 dB). It achieves higher Normalized Correlation (NC) after most attacks — including median filtering, Gaussian noise, resize, contrast adjustment, and more — while also being the only method secure against ambiguity attacks on both the watermark and the cover.
Will I receive a certificate for attending?
Yes. All registered participants who attend the live session will receive a digital certificate of attendance from the organizers, along with access to supplementary study materials related to the webinar topics.
Is the webinar really free?
Yes — this webinar is completely free to attend. Simply register using the form on the right, and you will receive joining instructions and reminders via email before the event.

Reserve Your Spot Today

May 10, 2026 · 12:00 p.m. Pacific Time · Online · Free

    Webinar Curriculum

    A structured journey through SVD-based color image watermarking — from fundamental concepts to advanced authentication mechanisms and real-world applications.

    Module 01
    Foundations of Digital Image Watermarking

    Overview of DIW types (spatial vs. transform domain, blind vs. non-blind). Key properties: robustness, imperceptibility, and embedding capacity. Components: watermark generation, embedding, extraction, and detection.

    Module 02
    SVD in Image Watermarking — and Its Flaw

    How Singular Value Decomposition works and why it is used in DIW. The fundamental false positive problem: how an attacker can extract a plausible fake watermark using substitute singular vectors of any image.

    Module 03
    The Proposed Technique — DWT + SVD + PC Embedding

    3-level DWT on the HSV Value plane. Computing Principal Components (PC = UW×SW) of two watermarks. Embedding PCs into H3 and V3 sub-band singular values with a tunable strength factor.

    Module 04
    Cover-Dependent Key Generation & Authentication

    How four keys (Key1–Key4) are generated from the cover and watermarks during embedding. Why knowing the keys alone is insufficient — the correct DWT coefficient combination is also required. Resistance to false cover claims demonstrated with NC tables.

    Module 05
    Robustness Against Image Processing Attacks

    Watermark survival across 20+ attacks: median filtering, Gaussian noise, salt & pepper noise, rotation, cropping, motion blur, resize, histogram equalization, and contrast adjustment. PSNR vs. sf trade-off analysis.

    Module 06
    Comparison with Existing Methods

    Side-by-side comparison with Ansari & Pant (2017) and Shivani & Senapati (2018). The proposed method outperforms existing work on most attacks at equivalent PSNR. Discussion of the independent sf(H) / sf(V) tuning strategy.

    Module 07
    Real-World Applications & Future Directions

    Copyright protection, medical image integrity (tamper localization), digital forensics, and secure multimedia communication. Future work: optimizing strength factors automatically based on cover and watermark characteristics.

    Module 08
    Q&A, Discussion & Certificate

    Open Q&A session with Dr. Neha Singh. Discussion of research opportunities in image security. All participants receive a certificate of attendance and access to supplementary materials.

    What Participants Say

    Don't Miss This Session

    Limited seats available. Register now and learn how to build watermarking systems that are genuinely secure against false positive attacks.

    Register Free — May 10, 2026

    Supporting UN Sustainable Development Goals

    This webinar advances digital security, research, and education — contributing directly to three key Sustainable Development Goals.

    Primary Focus

    SDG 4: Quality Education

    Key Target
    Target 4.7: Ensure learners acquire knowledge and skills needed to promote sustainable development through technical education.

    Provides researchers and students with practical, research-backed knowledge of SVD watermarking, authentication mechanisms, and false positive reduction — advancing technical education in cybersecurity and signal processing.

    SDG 3: Good Health & Well-Being

    Key Target
    Target 3.4: Support health through integrity of medical data systems and trusted diagnostic infrastructure.

    Robust watermarking of medical images (MRI, CT scans, X-rays) prevents tampering and ensures diagnostic accuracy — a direct application of the techniques taught in this webinar, supporting safer patient care.

    SDG 9: Industry, Innovation & Infrastructure

    Key Target
    Target 9.5: Enhance scientific research and innovation capabilities in developing and developed countries.

    Promotes advanced research in secure digital watermarking, contributing to robust multimedia infrastructure, innovation in signal processing, and stronger cybersecurity frameworks for digital economies worldwide.

    How This Webinar Contributes to Global Goals

    Advancing Secure Digital Ownership

    Teaching SVD + DWT + Principal Component embedding equips practitioners to build copyright protection systems that are mathematically resistant to false ownership claims.

    Ensuring Authenticity & Trust

    Cover-dependent key generation ensures that only the genuine owner can authenticate the watermark — closing a long-standing vulnerability in standard SVD watermarking systems.

    Supporting Cybersecurity Innovation

    Empowering engineers and researchers with validated, published techniques that outperform existing methods — driving innovation in digital forensics, medical imaging, and secure multimedia communications.