I wish to produce training images of text from a specific font with a font
size (40 pts) specific to my target domain. So far, training on the
standard 12 point size produced poor results for differentiating complex
kanji characters.
I have identified the relevant line of code in text2image.cpp, lines 89-90:
// Size of text (in points).
INT_PARAM_FLAG(ptsize, 12, "Size of printed text");
Should I directly change the value to 40 here in text2image.cpp or in
tesstrain_utils.sh lines 229-233:
local common_args="--fontconfig_tmpdir=${FONT_CONFIG_CACHE}"
common_args+=" --fonts_dir=${FONTS_DIR} --strip_unrenderable_words"
common_args+=" --leading=${LEADING}"
common_args+=" --char_spacing=${CHAR_SPACING} --exposure=${EXPOSURE}"
common_args+=" --outputbase=${outbase} --max_pages=${MAX_PAGES}"
such as adding a --ptsize 40 argument after --leading?
--
You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/tesseract-ocr.
To view this discussion on the web visit
https://groups.google.com/d/msgid/tesseract-ocr/3f62decb-0053-4382-9dda-0d8e96c52e56%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
