Tesseract

Tesseract is an OCR engine for developers and document-processing workflows that need to extract text from images. It includes a command-line tool and an embeddable library, but doesn't include a graphical app.

• Image text recognition: Read text from PNG, JPEG and TIFF images, with UTF-8 support and recognition across many languages.
• Searchable documents: Produce plain text or searchable PDFs, including PDFs containing only an invisible text layer.
• Structured results: Export hOCR HTML, TSV, ALTO and PAGE output for software that needs structured recognition results.
• Application integration: Add OCR to your own software through the libtesseract C or C++ API, or use the command-line tool in scripted workflows.
• Language training: Train recognition models for additional languages; the engine supports neural line recognition and legacy character recognition, which requires compatible trained language data.

Tesseract is open source under Apache 2.0. Recognition quality depends on the input image, so poor scans may need cleanup before processing. It accepts images rather than PDF documents as input.