OCR and Text Search App 📸🔍

This repository contains a Streamlit-based application that performs Optical Character Recognition (OCR) on uploaded images and enables users to search and highlight specific text within the extracted results.

Features ✨

• OCR: Extracts text from uploaded images using the ucaslcl/GOT-OCR2_0 model. 📝
• Search: Allows users to search for specific terms in the OCR results, highlighting matches. 🔎
• Interactive UI: Provides an easy-to-use interface for image uploading, OCR processing, and text search. 🖥️
• Dynamic Text Highlighting: Highlights search terms in the OCR results using a distinct color. 🎨

Installation and Setup ⚙️

Prerequisites 📋

• Python 3.8 or later 🐍
• Access to the Hugging Face model repository with a valid authentication token 🔑

Steps to Install 🚀

1. Clone the Repository

   git clone https://github.com/akhil-varsh/ocr-key-word-search.git
   cd ocr-key-word-search

2. Install Dependencies
   Install the required Python packages listed in requirements.txt:

   pip install -r requirements.txt

3. Set Your Hugging Face Token
   Replace REPLACE_YOUR_HF_TOKEN in the code with your Hugging Face API token.
   If you don’t have one, create an account at Hugging Face and generate an API token. 📝

4. Run the App
   Launch the application with the following command:

   streamlit run app.py

Usage 🛠️

1. Upload an Image

   • Click the "Choose an image..." button to upload a .jpg, .jpeg, or .png file. 🖼️

2. Run OCR

   • Click the "Run OCR" button to extract text from the uploaded image. ⚡

3. Search Text

   • Enter a term in the search input field to highlight matches in the OCR result. 🔍
   • View all matched lines under "Search Results." 📑

4. View OCR Output

   • Check the OCR results in the text area displayed below the "OCR Result" section. 💬

File Structure 📂

ocr-key-search-app/
├── app.py              # Main Streamlit application code 💻
├── requirements.txt    # Python dependencies 📦
├── README.md           # Documentation for the project 📄

Dependencies 📜

• Streamlit: For building the web interface. 🖥️
• Transformers: For accessing and running the GOT-OCR2_0 model. 🤖
• Pillow: For image handling. 🖼️
• NumPy: For image processing. 🔢
• PyTorch: Required by the OCR model for inference. ⚡

Acknowledgments 🙏

• Hugging Face: Providing the ucaslcl/GOT-OCR2_0 model. 🤝
• Streamlit: For making interactive web apps simple to build. 🏗️

Future Improvements 🚀

• Add support for multi-language OCR. 🌍
• Include an option to download the OCR results as a text file. 💾
• Improve search functionality with regex support. 🔠
• Optimize performance for large images. 🚀

License 📜

This project is licensed under the MIT License - see the LICENSE file for details. 🔓

Feel free to contribute or raise issues to improve the app! 🌟

---

Made with ❤️ by Akhil