ocr-backend-pluginlisted

Implement custom OCR backends 1. Create struct implementing OcrBackend 2. Initialize OCR model/engine 3. Implement async process_image(): a. Accept image bytes b. Extract language from config c. Validate image format d. Apply preprocessing if needed e. Run OCR engine f. Parse results g. Return ExtractionResult 4. Implement supported_languages() 5. Declare capabilities() 6. Handle async execution 7. Test OCR accuracy 8. Benchmark performance