🔍 Document Forgery Detection

An AI-powered full-stack web application that detects forged or tampered documents using advanced image forensic techniques and deep learning.

Features · Tech Stack · Getting Started · How It Works · Project Structure · Screenshots · Contributing

---

📌 Overview

Document forgery is a growing threat in legal, financial, and government sectors. This project provides an automated solution that analyses uploaded document images and flags potential manipulations — including splicing, copy-move, and region retouching — using a combination of:

• Error Level Analysis (ELA)
• Copy-Move Forgery Detection
• CNN-based Deep Learning Classification
• Weighted Decision Fusion

Results are served through a clean, responsive web interface where users can upload a document and get an instant verdict with visual heatmap overlays.

---

✨ Features

• 📄 Multi-format support — Accepts JPEG, PNG, and PDF-rendered document images
• 🧠 AI-powered detection — Fine-tuned EfficientNet-B0 for high-accuracy classification
• 🗺️ Visual forensics — ELA heatmaps and copy-move region masks overlaid on the original document
• ⚡ Fast inference — End-to-end analysis in under 2 seconds
• 📊 Confidence scoring — Probability score alongside the Authentic / Forged verdict
• 📥 Downloadable reports — Export a PDF summary of the analysis
• 📱 Responsive UI — Works seamlessly on desktop, tablet, and mobile

---

🛠 Tech Stack

Layer	Technology
Frontend	Next.js 14, TypeScript, Tailwind CSS
Backend / API	FastAPI (Python)
ML / Image Processing	Python, OpenCV, NumPy, Pillow, TensorFlow / PyTorch
Deep Learning Model	EfficientNet-B0 (fine-tuned)
Storage	Local filesystem / AWS S3
Containerisation	Docker

---

🚀 Getting Started

Prerequisites

• Node.js ≥ 18.x
• Python ≥ 3.10
• npm or yarn
• pip

---

1. Clone the Repository

git clone https://github.com/RaghavChandak112404/Document_Forgery_Detection.git
cd Document_Forgery_Detection

---

2. Backend Setup (Python)

cd backend

# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate        # macOS / Linux
venv\Scripts\activate           # Windows

# Install dependencies
pip install -r requirements.txt

# Start the FastAPI server
uvicorn api.main:app --reload --port 8000

The API will be available at http://localhost:8000.
Interactive docs at http://localhost:8000/docs.

---

3. Frontend Setup (Next.js)

cd Document_Forgery_Detection   # frontend directory

# Install dependencies
npm install

# Set environment variable
cp .env.example .env.local
# Set NEXT_PUBLIC_API_URL=http://localhost:8000 in .env.local

# Start the dev server
npm run dev

The app will be available at http://localhost:3000.

---

4. Run with Docker (Optional)

# From the project root
docker-compose up --build

This starts both the frontend and backend in isolated containers.

---

🔬 How It Works

The detection pipeline runs in four stages:

Document Upload
      │
      ▼
 Preprocessing
 (resize, normalise, estimate JPEG quality)
      │
      ├──────────────────────────────────┐──────────────────────────────┐
      ▼                                  ▼                              ▼
Error Level Analysis (ELA)     Copy-Move Detection            CNN Classification
  Detects spliced regions        Finds duplicated areas        EfficientNet-B0
  via JPEG compression           within the document           probability score
  anomalies
      │                                  │                              │
      └──────────────────────────────────┴──────────────────────────────┘
                                         │
                                         ▼
                              Weighted Decision Fusion
                         S = w₁·ELA + w₂·CopyMove + w₃·CNN
                                         │
                                         ▼
                             Verdict: Authentic / Forged
                             + Confidence Score + Overlays

Detection Methods

Method	Detects	How
Error Level Analysis	Splicing, region replacement	JPEG re-compression differential
Copy-Move Detection	Duplicated regions	ORB keypoints + RANSAC + DCT block comparison
CNN Classification	General manipulations	Fine-tuned EfficientNet-B0

---

📁 Project Structure

Document_Forgery_Detection/
│
├── Document_Forgery_Detection/     # Next.js frontend (TypeScript)
│   ├── app/                        # App router pages
│   │   ├── page.tsx                # Home / Upload page
│   │   ├── results/page.tsx        # Results visualisation page
│   │   └── history/page.tsx        # Past analyses
│   ├── components/                 # Reusable UI components
│   ├── lib/                        # API client, utilities
│   ├── public/                     # Static assets
│   └── package.json
│
├── backend/                        # Python ML service
│   ├── api/
│   │   ├── main.py                 # FastAPI entry point
│   │   ├── routers/detect.py       # /detect endpoint
│   │   └── schemas.py              # Pydantic models
│   ├── ml/
│   │   ├── ela.py                  # Error Level Analysis
│   │   ├── copy_move.py            # Copy-Move detection
│   │   ├── cnn_model.py            # EfficientNet inference
│   │   └── fusion.py               # Decision fusion
│   ├── utils/
│   │   └── image_utils.py          # Preprocessing helpers
│   └── requirements.txt
│
├── package-lock.json
└── README.md

---

📊 Performance

Method	Accuracy	Precision	Recall	F1	AUC
ELA only	74.2%	0.71	0.79	0.75	0.81
Copy-Move only	69.8%	0.68	0.73	0.70	0.76
CNN only	88.6%	0.87	0.90	0.88	0.94
Ensemble (Proposed)	91.3%	0.90	0.93	0.91	0.96

Evaluated on a 20% holdout test set combining CASIA v2.0, Columbia, and custom document data.

---

📸 Screenshots

Add screenshots of your UI here by placing images in a /screenshots folder and updating the paths below.

Upload Page	Results Page

---

🌐 API Reference

POST /detect

Analyse a document image for forgery.

Request

Content-Type: multipart/form-data

file: <document image>

Response

{
  "verdict": "forged",
  "score": 0.87,
  "ela_map": "<base64 encoded heatmap>",
  "cm_mask": "<base64 encoded mask>",
  "processing_time_ms": 812
}

---

🤝 Contributing

Contributions are welcome! To get started:

1. Fork the repository
2. Create a feature branch: git checkout -b feature/your-feature-name
3. Commit your changes: git commit -m "feat: add your feature"
4. Push to your branch: git push origin feature/your-feature-name
5. Open a Pull Request

---

👤 Author

Raghav Chandak

---

_{If you found this project helpful, consider giving it a ⭐}