The Fake Payment Screenshot Detector is an AI-powered web-based tool designed to identify forged or tampered payment transaction screenshots. It uses image forensics, machine learning, and metadata analysis to verify the authenticity of payment proof images submitted during transactions.
Whether you're a business, freelancer, or platform that accepts digital payments, this tool helps protect against scams involving edited screenshots from apps like Google Pay, PhonePe, Paytm, or bank portals.
-
✅ Image Forensics (ELA & Edge Detection)
Detects manipulations using Error Level Analysis (ELA), edge inconsistencies, and compression anomalies. -
🧠 AI-Based Classification
Trained ML model distinguishes between real and fake transaction images with high accuracy. -
🕵️ OCR Verification
Extracts text using Tesseract OCR to check if the content (like transaction ID, amount, etc.) matches expected patterns. -
📸 Metadata Analysis
Examines image EXIF data to determine the source (camera, WhatsApp, screenshots, etc.). -
🖼️ Image Preview + Drag & Drop UI
Users can preview uploaded images with a clean, responsive drag-and-drop interface. -
📄 PDF Report Generation
Generates downloadable forensic reports summarizing the analysis results. -
🌓 Dark Mode Support
Fully styled for light and dark themes to enhance usability.
| Layer | Technology Used |
|---|---|
| Frontend | React.js, Bootstrap |
| Backend | Flask / Express.js |
| AI/ML | TensorFlow, OpenCV, scikit-learn |
| OCR | Tesseract |
| Database | MongoDB (for user tracking, history) |
| Extras | Pillow, PDFKit/ReportLab, EXIFTools |
- User uploads a payment screenshot
- Backend processes the image:
- Performs ELA & Edge Detection
- Extracts text using OCR
- Checks metadata for social media signs (e.g. WhatsApp compression)
- Uses an AI model to classify authenticity
- Results are displayed with visual evidence
- PDF report is generated and optionally sent via email
fake-payment-detector/ │ ├── client/ # React frontend │ ├── public/ │ └── src/ │ ├── components/ │ ├── pages/ │ └── App.js │ ├── server/ # Flask/Express backend │ ├── models/ │ ├── utils/ │ ├── routes/ │ └── app.py / index.js │ ├── ml_model/ # ML training and saved model │ ├── training.ipynb │ └── model.h5 / model.pkl │ ├── README.md └── requirements.txt / package.json
yaml Copy Edit
- 📥 WhatsApp/Telegram Image Source Detector
- 🎥 Live Camera Upload Support
- 📧 Report Sharing via Email/WhatsApp
- 🧾 Real vs Fake Comparison View
- 🔐 User Login & History Tracking
- ⛔ Deepfake Nude Image Detection Add-on (in progress)
- Freelancers accepting UPI payments
- Online sellers verifying buyer payments
- Businesses preventing digital transaction scams
- Forensics and cybercrime investigators
Developed by: Rashmi Kumari
Feel free to reach out via Email: rashmi8shahi@gmail.com for collaboration or suggestions!
This project is licensed under the MIT License — feel free to use and contribute.
Found a bug? Want to add a feature? Contributions are welcome!
Please fork the repository and submit a pull request.