bottle-ocr-lib

🏥 BottleOCR Library

Professional OCR and AI-powered prescription extraction library with secure license validation and local processing capabilities.

✨ Features

• 🔍 High-Accuracy OCR - Advanced PaddleOCR text extraction
• 🤖 AI-Powered Analysis - OpenAI GPT-4 prescription data extraction
• 🔐 Secure License System - One-time server validation, then local processing
• 🖼️ Multi-Format Support - JPEG, PNG, TIFF, BMP, PDF, numpy arrays, PIL Images
• ⚡ Batch Processing - Process multiple images efficiently
• 💾 Smart Caching - Offline operation after initial validation
• 🛠️ Easy Integration - Simple API with comprehensive documentation
• 📊 Structured Output - 18+ prescription fields extracted automatically

🚀 Quick Start

Installation

pip install bottle-ocr-lib

Basic Usage

from bottle_ocr_lib import BottleOCR

# Initialize with your API key (validates once, then works offline)
ocr = BottleOCR(api_key="your-api-key-here")

# Process prescription bottle images
result = ocr.process_single_image("prescription_bottle.jpg")

# Access extracted prescription data
prescription = result['prescription']
print(f"💊 Medication: {prescription['medication_name']}")
print(f"💉 Dosage: {prescription['dosage']}")
print(f"👤 Patient: {prescription['patient_name']}")
print(f"🏥 Pharmacy: {prescription['pharmacy_name']}")

Batch Processing

# Process multiple images at once
results = ocr.process_images([
    "bottle_front.jpg",
    "bottle_back.jpg", 
    "label_close_up.png"
])

for i, result in enumerate(results['images']):
    if result['status'] == 'success':
        prescription = result['prescription']
        print(f"Image {i+1}: {prescription['medication_name']}")

🔐 License System

BottleOCR uses a secure validation system enabling complete local processing:

• Initial Validation: Your API key validates with our server (one-time only)
• Encoded Key Delivery: Server provides encrypted OpenAI API key
• Local Processing: All subsequent operations run offline on your machine
• Smart Caching: No repeated server communication required

# First run: Server validation + local caching
ocr = BottleOCR(api_key="your-key")  # ✅ Online validation

# All future runs: Instant startup from cache  
ocr = BottleOCR(api_key="your-key")  # ✅ Offline, instant
results = ocr.process_images(images)  # ✅ 100% local processing

📊 Extracted Prescription Data

The library extracts comprehensive prescription information:

Field	Description	Example
medication_name	Drug name	"Amoxicillin"
dosage	Strength/amount	"500mg"
quantity_dispensed	Amount given	"30 capsules"
patient_name	Patient name	"John Doe"
prescriber_name	Doctor name	"Dr. Smith"
pharmacy_name	Pharmacy name	"Main St Pharmacy"
prescription_date	Fill date	"2024-10-25"
expiration_date	Expiry date	"2025-10-25"
directions_for_use	Instructions	"Take twice daily"
refills_remaining	Refills left	"2"
rx_number	Prescription #	"RX7654321"
ndc_number	NDC code	"12345-678-90"
lot_number	Lot number	"ABC123"
manufacturer	Drug maker	"Generic Co"
warning_labels	Warnings	"May cause drowsiness"
storage_instructions	Storage	"Store at room temp"
dosage_form	Form type	"Capsule"
description_of_pill	Appearance	"Blue oval tablet"

🖥️ Command Line Interface

# Process single image
bottle-ocr process image.jpg --api-key your-key

# Process multiple images
bottle-ocr batch *.jpg --output results.json --api-key your-key

# Get account information
bottle-ocr info --api-key your-key

# Validate images without processing
bottle-ocr validate image1.jpg image2.jpg --api-key your-key

⚙️ Configuration

Environment Variables

export BOTTLEOCR_API_KEY="your-api-key"

Custom Configuration

config = {
    "ocr": {
        "language": "en",
        "confidence_threshold": 0.8,
        "use_gpu": True
    },
    "extraction": {
        "model": "gpt-4",
        "temperature": 0.1
    }
}

ocr = BottleOCR(api_key="your-key", config=config)

📝 Complete Example

from bottle_ocr_lib import BottleOCR
from bottle_ocr_lib.utils.exceptions import AuthenticationError, ValidationError

try:
    # Initialize (validates license once)
    ocr = BottleOCR(api_key="your-api-key")
    
    # Process images (works offline after validation)
    results = ocr.process_images([
        "prescription1.jpg",
        "bottle_label.png"
    ])
    
    # Extract prescription information
    for i, result in enumerate(results['images']):
        if result['status'] == 'success':
            p = result['prescription']
            print(f"""
Image {i+1}:
  Medication: {p['medication_name']}
  Dosage: {p['dosage']}
  Patient: {p['patient_name']}
  Instructions: {p['directions_for_use']}
  Refills: {p['refills_remaining']}
            """)
        else:
            print(f"Image {i+1} failed: {result['error']}")
            
except AuthenticationError as e:
    print(f"❌ License validation failed: {e}")
except ValidationError as e:
    print(f"❌ Invalid input: {e}")
except Exception as e:
    print(f"❌ Unexpected error: {e}")

🔧 Requirements

• Python: 3.8 or higher
• Platform: Windows, macOS, Linux
• Dependencies: Automatically installed
  • PaddleOCR >= 2.7.0
  • OpenAI >= 1.0.0
  • OpenCV >= 4.5.0
  • Pillow >= 8.0.0
  • PyYAML >= 5.4.0

🆘 Getting an API Key

• Sign up: Visit bottleocr.com to create an account
• Choose Plan: Select the subscription that fits your needs
• Get Key: Receive your API key via email after signup
• Start Processing: Use your key to process prescription images immediately

📖 Documentation & Support

• 📚 Examples: See examples/ directory for complete usage patterns
• 🐛 Issues: GitHub Issues
• 💬 Support: [e-mail]
• 📖 API Docs: Comprehensive docstrings in all methods

📄 License

MIT License - see LICENSE file for details.

🏆 Why Choose BottleOCR?

• ✅ Production Ready - Battle-tested accuracy and reliability
• ✅ Privacy First - Your data stays on your machine after validation
• ✅ Developer Friendly - Simple API, great documentation, quick setup
• ✅ Cost Effective - Pay once, process locally forever
• ✅ Scalable - Handle single images or large batch processing
• ✅ Secure - Encrypted license validation with local operation

Ready to extract prescription data professionally? Get your API key today! 🚀