Extracts higher-quality OCR text from poorly scanned PDFs and images (PDF, PNG, JPG, BMP, TIFF) by fixing common scan errors: crops to estimated text regions, deskews tilted pages, and optimizes images before running Tesseract OCR; supports bulk processing (splitting multi-page PDFs) and language-code hints for accuracy.