Home

Vision Text Extractor Wiki

Welcome to the Vision Text Extractor wiki! This comprehensive guide will help you get the most out of this multi-provider OCR tool.

🚀 Quick Navigation

Getting Started

• Installation Guide - Complete setup instructions
• Quick Start Tutorial - Get running in 5 minutes
• Configuration - Environment setup and API keys

Usage Guides

• Basic Usage - Command-line interface basics
• Advanced Features - Custom prompts, batch processing
• Provider Comparison - Choosing the right AI model

Tutorials

• Document Processing - Business documents, forms, receipts
• Recipe Extraction - Cooking and food industry use cases
• Academic Research - Educational and research applications

Technical Reference

• API Reference - Function documentation
• Pixi Tasks - Complete task reference
• Troubleshooting - Common issues and solutions

Development

• Contributing - How to contribute to the project
• Architecture - Technical architecture overview
• Adding New Providers - Extend with more AI models

🌟 What is Vision Text Extractor?

Vision Text Extractor is a powerful, multi-provider OCR (Optical Character Recognition) tool that uses state-of-the-art vision-language models to extract text from images and documents.

Key Features

• 3 AI Providers: Choose from Hugging Face SmolVLM (local), Ollama LLaVA (local), or OpenAI GPT-4o (cloud)
• Flexible Input: Support for local files and web URLs
• Custom Prompts: Extract specific information with targeted prompts
• Professional CLI: Built with Typer for excellent user experience
• Easy Setup: Managed dependencies with Pixi

Use Cases

• 📄 Business Documents: Contracts, invoices, forms
• 🍽️ Food Industry: Recipes, menus, nutrition labels
• 💰 Finance: Receipts, bank statements, tax documents
• 📚 Education: Homework, research papers, lecture notes
• 🏥 Healthcare: Prescriptions, lab results, medical forms
• 🏠 Real Estate: Property listings, lease agreements

🎯 Quick Example

# Install and setup
git clone https://github.com/udit-asopa/vision-text-extractor.git
cd vision-text-extractor
pixi install
pixi run setup

# Extract text from an image
pixi run demo-ocr-huggingface
python main.py path/to/your/image.jpg

# Custom extraction
python main.py receipt.jpg --prompt "Extract total amount and date"

📚 Learning Path

New Users: Start with Installation Guide → Quick Start Tutorial

Power Users: Jump to Advanced Features → API Reference

Developers: Check out Architecture → Contributing

---

Last updated: October 2025