使用 Homebrew, apt, dnf, MacPorts, Nix, pacman, zypper 安装 ocrad

项目链接

• 主页

命令和别名

• ocrad

项目历史

The GNU Ocrad manual carries copyright years 2003-2024 for Antonio Diaz Diaz, and the Free Software Directory identifies it as a GNU package. GNU’s FTP archive shows public release tarballs for the 0.14 through 0.17 series in 2006 and 2007, while the manual documents version 0.29 dated 18 January 2024.

The manual characterizes Ocrad as mainly a research project with ad hoc algorithms that may change as the author gains OCR understanding. That framing explains both the project’s small scope and its historical importance: it is a free-software OCR engine built around explicit image-processing and character-recognition code, not a wrapper around a larger OCR system.

采用历史

Ocrad entered package repositories as a small GNU utility and library for OCR workflows before Tesseract dominated most open-source OCR packaging. Homebrew, Debian/Ubuntu, Fedora, MacPorts, Nix, Arch, and openSUSE-family metadata list packages for it or its development library, keeping it available for scripts and programs that need a lightweight OCR backend.

使用方式

Users invoke `ocrad` on PNG, PBM, PGM, or PPM images, optionally enabling layout analysis, scaling, thresholding, rotation, character-set selection, postprocessing filters, and UTF-8 output. The manual also documents conversion pipelines from PostScript, PDF, TIFF, JPEG, and compressed PNM through Ghostscript or Netpbm-style tools before recognition.

As a library, Ocrad exposes functions for loading images, setting output format and thresholds, running recognition, and retrieving block, line, and character results. Its OCR results file format gives downstream programs access to text blocks, line positions, character boxes, guesses, and confidence values.

为什么软件包爱好者会关心

For package metadata, Ocrad is a good example of a GNU-era single-purpose C++ utility that is both an executable and a library. It connects classic Unix image pipelines, OCR experimentation, GPL packaging, and downstream development packages such as Debian and Ubuntu’s `libocrad-dev`.

时间线

• 2003: the manual’s copyright range begins, matching Ocrad’s early GNU-era development period.
• 2006-2007: GNU FTP archives list 0.14, 0.15, 0.16, and 0.17 release tarballs.
• 2022: the Free Software Directory listed 0.28 as a stable release dated 17 January 2022.
• 2024-01-18: the GNU manual documents version 0.29.

Related projects

• Ocrad belongs with GNU image/text-processing utilities, Netpbm and Ghostscript conversion pipelines, and OCR front ends or callers that can use a command-line OCR backend.

来源

• https://directory.fsf.org/wiki/Ocrad
• https://ftp.gnu.org/gnu/ocrad/
• https://savannah.gnu.org/projects/ocrad/
• https://www.gnu.org/software/ocrad/manual/ocrad_manual.html
• source_facts.executables
• source_facts.package-manager