The release of the optical character recognition system Tesseract 5.3.4 has been published, supporting UTF-8 character recognition and texts in over 100 languages, including Russian, Kazakh, Belarusian, and Ukrainian. The output can be saved as plain text as well as in HTML (hOCR), ALTO (XML), PDF, and TSV formats. Originally, the system was developed between 1985 and 1995 in the laboratory of Hewlett Packard, and in 2005, the code was released under the Apache license and has since been further developed with the involvement of Google employees. The project's source texts are distributed under the Apache 2.0 license.
Tesseract includes a console utility and the libtesseract library for embedding text recognition features into other applications. Notable third-party GUI interfaces supporting Tesseract include gImageReader, VietOCR, and YAGF. It offers two recognition engines: the classic one, which recognizes text by matching individual character templates, and a new one based on a recurrent neural network (LSTM) machine learning system, optimized for recognizing entire lines of text and significantly improving accuracy. Pre-trained models are available for 123 languages. Performance optimization modules using OpenMP and SIMD instructions such as AVX2, AVX, AVX512F, NEON, or SSE4.1 are provided.
Key Improvements:
- Image recognition via URL has been enhanced with file upload using the libcurl library. The User-Agent header is set during the upload. A new parameter curl_cookiefile has been added to utilize a cookie file.
- In server TCP has been employed as the priority protocol for ScrollView.
- When using the 'combine_tessdata -d' command, output is now directed to stdout instead of stderr.
- Issues with building using autoconf and clang have been resolved.
Source: opennet.ru
