VisionScript

A cutting-edge text extraction and summarization solution built with React, TypeScript, and Python, featuring a modern tech stack and intelligent document processing capabilities. Transform your physical documents into digital format with ease and precision.

🚀 Recent Updates

• AI Text Summarization - Added intelligent text summarization with OpenRouter Mistral integration
• PDF Support - Upload and process PDF documents for text extraction and summarization
• Smart Response Cleaning - AI responses automatically formatted for human readability
• Multilanguage Support - Added support for 13+ languages with seamless language switching
• UI Improvements - Enhanced user interface for better user experience
• Backend Optimization - Improved text detection accuracy and processing speed

🌟 Key Features

📸 Smart Document Capture

• Drag-and-drop file upload support
• Real-time camera-based document scanning
• Intelligent frame capture and processing
• Support for multiple image formats (JPG, PNG, BMP)
• Advanced image preprocessing for optimal results

🔍 Powerful OCR Processing

• High-accuracy text extraction using Tesseract OCR
• Smart document layout analysis
• Automatic document deskewing and orientation correction
• Noise reduction and image enhancement
• Confidence-based text filtering
• Multi-language support for text recognition in 13+ languages

📄 Advanced Export Options

• Multiple export formats:
  • Text files (TXT) with UTF-8 encoding
  • Word documents (DOCX) with formatting preservation
  • Excel spreadsheets (XLSX) for tabular data
• Smart paragraph and layout detection
• Header/footer identification
• Table structure recognition

🎯 Real-Time Processing

• Live camera feed processing
• Instant text extraction feedback
• Progress indicators and status updates
• Performance optimization for smooth operation
• Frame rate and resolution control

🛡️ Robust Error Handling

• Comprehensive input validation
• Detailed error messaging
• Automatic retry mechanisms
• Fallback strategies
• Secure file handling

🤖 AI-Powered Text Summarization

• Smart Summarization Options:
  • Local Smart Summary - Advanced algorithms for intelligent text condensation
  • AI Summary (OpenRouter Mistral) - High-quality AI-powered summarization
• Multiple Summary Formats:
  • Paragraph summaries for natural reading
  • Bullet point summaries for quick scanning
  • Key phrase extraction for topic identification
• Customizable Length Control:
  • Short (25%) - Quick overview in 2-3 sentences
  • Medium (50%) - Balanced summary in 3-4 sentences
  • Long (75%) - Detailed summary in 5-7 sentences
• PDF Support - Upload and summarize PDF documents directly
• Rate Limit Management - Real-time API usage monitoring with cooldown display
• Clean Output - AI responses automatically cleaned for human readability

🌐 Multi-Language Support

• Support for 13+ languages including:
  • English, French, Spanish, German
  • Italian, Portuguese, Dutch
  • Chinese, Japanese, Korean
  • Russian, Arabic, Hindi
• Language-specific text recognition optimization
• Easy language selection via dropdown interface
• Real-time language switching

📁 Project Structure

VisionScript/
├── src/             # React + TypeScript frontend
│   ├── components/  # Reusable UI components
│   │   ├── ImageUpload.tsx       # Image upload component
│   │   ├── RealTimeDetection.tsx # Camera-based detection
│   │   ├── LanguageSelector.tsx  # Language selection component
│   │   └── TextSummarizer.tsx    # Text summarization component
│   └── styles/      # Tailwind CSS styling
├── backend/         # Python Flask backend
│   ├── main_test.py # Main application server
│   └── .env         # Environment configuration
├── docs/           # Documentation
└── node_modules/   # Dependencies

🚀 Getting Started

1. Clone the repository:

   git clone git@github.com:imadnan4/VisionScript.git

2. Navigate to the project directory:

   cd VisionScript

3. Install dependencies:

   npm install
   cd backend
   pip install -r requirements.txt

4. Start the development server:

   npm run dev

5. (Optional) Set up AI Text Summarization:

   • Get a free API key from OpenRouter
   • Add to backend/.env file:

     OPENROUTER_API_KEY=your_api_key_here
     

   • Restart the backend server to enable AI summarization

6. Using the multilanguage feature:

   • Look for the globe icon in the interface
   • Click to open the language dropdown
   • Select your preferred language for OCR processing
   • The system will automatically optimize text recognition for the selected language

7. Using the text summarizer:

   • Upload images/PDFs or use camera to extract text
   • Click "Summarize Text" button on extracted text
   • Or go directly to "Text Summarizer" from the home screen
   • Choose between Local Smart Summary or AI Summary (OpenRouter)
   • Select format (paragraph, bullets, key phrases) and length
   • Export summaries in TXT, DOCX, or Excel format

🛠️ Technical Stack

Frontend

• React 18+ with TypeScript
• Vite for fast development
• Tailwind CSS for modern styling
• Framer Motion for smooth animations
• React Context API for state management

Backend

• Python Flask for API endpoints
• Tesseract OCR for text extraction
• EasyOCR for multi-language support
• OpenCV for image processing
• python-docx for Word document generation
• PyPDF2 for PDF text extraction
• Text Summarization Stack:
  • SUMY library for local extractive summarization
  • NLTK for natural language processing
  • OpenRouter API integration for AI-powered summarization
  • Mistral AI model for high-quality text generation
  • Custom response cleaning algorithms

🔜 Upcoming Features

• Text Translation - Integrate Google Translate API for multi-language text translation
• Voice-to-Text - Add speech recognition for audio file transcription
• Smart Text Comparison - Compare multiple documents and highlight differences
• Text Analytics Dashboard - Word count, reading time, sentiment analysis, and readability scores
• Custom Summary Templates - User-defined summary formats (executive summary, meeting notes, etc.)
• Batch Processing - Process multiple images/PDFs in one operation
• Text Search & Highlight - Search within extracted text and highlight matches
• Export to More Formats - Add support for Markdown, HTML, and plain text exports
• OCR Confidence Scoring - Display confidence levels for extracted text with editing suggestions
• Dark/Light Theme Toggle - Customizable UI themes for better user experience

📝 License

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

👤 Author

• Adnan
• GitHub: @imadnan4

🤝 Contributing

Contributions, issues, and feature requests are welcome! Feel free to check the issues page.