After dozens of hurdles solved, this one (or two) finally stopped me...
So I'm trying to train Tesseract 3.04 for specific font and I've completed couple of .box files with over 800 characters each. Then I've created .tr files and now I'm ready to train Tesseract:
sudo ./tesstrain.sh --lang eng --langdata_dir /Users/vitaliy/Desktop/tess-training/langdata --tessdata_dir /Users/vitaliy/Desktop/tess-training/TIFs
This is an error (s) I'm getting:
=== Starting training for language 'eng'
mktemp: illegal option -- -
usage: mktemp [-d] [-q] [-t prefix] [-u] template ...
mktemp [-d] [-q] [-u] -t prefix
ERROR: text2image not found
Any advice helps! Thank you!
I'm not sure what does the command that you have described do, but as for the error, text2image not found, try the following
as it was described in this website if you search for the command above it says that these commands are for training tools, which clearly text2image is one of them