OCR-Based Transaction Processing with EasyOCR & PDF2Image

Project Overview

This project automates the extraction of critical transaction details—Terminal ID, STAN (System Trace Audit Number), and RRN (Reference Retrieval Number)—from financial receipts and statements using EasyOCR and PDF2Image.

Features

✅ Convert PDFs to images for OCR processing
✅ Extract structured text using EasyOCR
✅ Preprocess and clean extracted data to improve accuracy
✅ Use Regular Expressions (Regex) to retrieve key transaction details
✅ Print extracted text and transaction details for debugging
✅ Lightweight and efficient Python-based implementation

Technologies Used

• Python (OS, Regex, PIL, Logging)
• EasyOCR (Deep-learning-based OCR for text extraction)
• PDF2Image (Convert PDFs into images for OCR processing)
• Regular Expressions (Extract structured transaction details)

Installation

Prerequisites

Ensure you have Python installed. Then, install the required dependencies:

pip install easyocr pdf2image pillow

Running the Project

1. Place your PDF receipt in the project folder.
2. Run the script to extract transaction details:

   python utils.py <path_to_pdf>

3. View extracted text and transaction details in the terminal.

Project Screenshots

📌

Video Demo

🎥 Watch full demo on LinkendIn

Future Enhancements

• ✅ NLP-based text correction for improved accuracy
• ✅ Multi-language support for receipts from various regions
• ✅ Web interface for real-time transaction processing using Django

Contributing

Pull requests are welcome! If you’d like to contribute, please fork the repository and submit a PR.

---

🚀 Let’s connect! If you find this project useful, feel free to star ⭐ the repo and connect with me on LinkedIn.