tesseract

Why we included this project

Tesseract is the engine most people mean when they say OCR. It's been the open-source standard for well over a decade, and it ships as both a command-line tool and a C/C++ library, so you can use it in a quick script or embed it in a larger document-processing pipeline. The current engine uses an LSTM neural network for line recognition, with a legacy mode kept for compatibility, and it reads more than a hundred languages out of the box, plus you can train it on new ones. Output can be plain text, hOCR, PDF, TSV, ALTO, or PAGE, so it's easy to hand results to downstream search, indexing, or data-extraction systems. Teams doing document digitization, receipt parsing, or archival work get a dependable, well-documented foundation with bindings in many languages.