OCR MCP Service

"""Recognize text in an image using OCR engines.""" import sys import json from pathlib import Path # Add project root to path before importing scripts.common project_root = Path(__file__).parent.parent sys.path.insert(0, str(project_root)) # Setup script environment from scripts.common import setup_script setup_script() def recognize_with_engine(image_path, engine_type="paddleocr", include_analysis=True): """Recognize text using specified engine.""" from ocr_mcp_service.ocr_engine import OCREngineFactory from ocr_mcp_service.utils import validate_image # Validate image validate_image(image_path) # Get engine and recognize print(f"Loading {engine_type} engine...") engine = OCREngineFactory.get_engine(engine_type) print("Recognizing text...") result = engine.recognize_image(image_path) # Remove analysis if not needed if not include_analysis: result.analysis = None return result def main(): """Main function.""" import argparse parser = argparse.ArgumentParser(description="OCR image recognition") parser.add_argument("image_path", help="Path to image file") parser.add_argument( "--engine", choices=["paddleocr", "paddleocr_mcp", "deepseek"], default="paddleocr", help="OCR engine to use (default: paddleocr)" ) parser.add_argument( "--json", action="store_true", help="Output as JSON" ) parser.add_argument( "--no-analysis", action="store_true", help="Disable technical analysis in output" ) args = parser.parse_args() # Use absolute path image_path = str(Path(args.image_path).resolve()) if not Path(image_path).exists(): print(f"Error: Image file not found: {image_path}") sys.exit(1) print("=" * 60) print(f"OCR Recognition - {args.engine}") print("=" * 60) print(f"Image: {image_path}") print() try: include_analysis = not args.no_analysis result = recognize_with_engine(image_path, args.engine, include_analysis=include_analysis) if args.json: # Output as JSON output = result.to_dict() print(json.dumps(output, ensure_ascii=False, indent=2)) else: # Human-readable output print("\n" + "=" * 60) print("Recognition Result") print("=" * 60) print(f"Engine: {result.engine}") print(f"Processing Time: {result.processing_time:.2f}s") print(f"Confidence: {result.confidence:.2f}") print(f"Text Boxes: {len(result.boxes)}") print("\nRecognized Text:") print("-" * 60) # Use get_text_with_analysis() if analysis is enabled, otherwise just text if include_analysis and result.analysis: print(result.get_text_with_analysis()) else: print(result.text) print("-" * 60) if result.boxes: print(f"\nText Boxes ({len(result.boxes)}):") for i, box in enumerate(result.boxes[:20], 1): # Show first 20 boxes print(f" Box {i}: ({box.x1:.1f}, {box.y1:.1f}) -> ({box.x2:.1f}, {box.y2:.1f})") if len(result.boxes) > 20: print(f" ... and {len(result.boxes) - 20} more boxes") except Exception as e: print(f"\nError: {e}") import traceback traceback.print_exc() sys.exit(1) if __name__ == "__main__": main()