OCRmyPDF

OCRmyPDF adds an OCR text layer to scanned PDFs so they can be searched or copied, and its README explains the frustrations that led to it.

OCRmyPDF generates a searchable PDF and PDF/A file from a regular PDF, placing OCR text accurately below the image so copy and paste works. When possible it inserts the OCR information as a lossless operation and keeps the exact resolution of the original embedded images. It also validates input and output files and, when requested, deskews or cleans the image before doing the OCR.

The author says they searched the web for a free command line tool to OCR PDFs and found many that fell short: misplaced text, missing accents and multilingual characters, changed resolutions, absurdly large files, crashes, invalid output, and none producing PDF/A at all. That list of failures is the motivation section.

The README says the tool distributes work across all available CPU cores, uses the Tesseract OCR engine for more than 100 languages, and scales to files with thousands of pages. It adds that it keeps private data private and has been used on millions of PDFs.

A plugin interface lets the OCR engine be replaced, with AppleOCR, EasyOCR, and PaddleOCR named as known plugins, and paperless-ngx integrates it into a document management system. Installation covers Linux, Windows, macOS, and FreeBSD, plus Docker images for x64 and ARM, and the software is licensed under MPL-2.0.

The README does the talking here, from the feature list to the licensing. Nothing in this review claims the tool was run or that its performance figures were checked.

• Source 1 · Project repository
• Source 2 · README