1SHAPECLUSTERING(1) 2================== 3:doctype: manpage 4 5NAME 6---- 7shapeclustering - shape clustering training for Tesseract 8 9SYNOPSIS 10-------- 11shapeclustering -D 'output_dir' 12 -U 'unicharset' -O 'mfunicharset' 13 -F 'font_props' -X 'xheights' 14 'FILE'... 15 16DESCRIPTION 17----------- 18shapeclustering(1) takes extracted feature .tr files (generated by 19tesseract(1) run in a special mode from box files) and produces a 20file *shapetable* and an enhanced unicharset. This program is still 21experimental, and is not required (yet) for training Tesseract. 22 23OPTIONS 24------- 25-U 'FILE':: 26 The unicharset generated by unicharset_extractor(1). 27 28-D 'dir':: 29 Directory to write output files to. 30 31-F 'font_properties_file':: 32 (Input) font properties file, where each line is of the following form, where each field other than the font name is 0 or 1: 33 34 'font_name' 'italic' 'bold' 'fixed_pitch' 'serif' 'fraktur' 35 36-X 'xheights_file':: 37 (Input) x heights file, each line is of the following form, where xheight is calculated as the pixel x height of a character drawn at 32pt on 300 dpi. [ That is, if base x height + ascenders + descenders = 133, how much is x height? ] 38 39 'font_name' 'xheight' 40 41-O 'FILE':: 42 The output unicharset that will be given to combine_tessdata(1). 43 44SEE ALSO 45-------- 46tesseract(1), cntraining(1), unicharset_extractor(1), combine_tessdata(1), 47unicharset(5) 48 49<https://tesseract-ocr.github.io/tessdoc/Training-Tesseract.html> 50 51COPYING 52------- 53Copyright \(C) Google, 2011 54Licensed under the Apache License, Version 2.0 55 56AUTHOR 57------ 58The Tesseract OCR engine was written by Ray Smith and his research groups 59at Hewlett Packard (1985-1995) and Google (2006-present). 60