Linux (Free) Tools for Optical Character Recognition
Free Linux tools
Demonstration on goocr
Interface pick you your file using browse
The example was a book from 1982 in pgm format
generally you can read the interpreted text but there were some errors with l's and b's and similar items.
ocrad -o (output) (Input)
Demonstration on OCRAD
didn't translate as well as gocr
Tesseract-OCR (Google Took over after 2007)
python3 /usrbin/pytesseract -l eng input output file
using tiff instead of pgm
tesseract-ocr was the best of the 3.
Images were massaged using GIMP to get text ready for OCR
Okular is also used to select area's and save them as inmates.
Examples from a scan from a really old print out from an impact printer did way worse than the scan from a book.
Font is also a factor on how well OCR does.
New versions of the above may make a difference on how well it worked.
Other options may help as well not all of those were explored in this presentation
Devon showed us
Http screen sharing Android 11 (lineage OS) He had an older phone so was using lineage to get a newer version of Android.
Android app (screen stream)
Program streams to a web page.
Sound will not stream sound from the Phone.
a tiny bit of a delay
showed off an android game using a game controller game was a bit delayed
This ends our meeting for the evening.