Verwende Python3 als Standard:
sudo apt-get install python-is-python3
Evtl python2 entfernen (noch nicht gemacht): https://askubuntu.com/questions/1242702/how-to-remove-python-2-from-ubuntu-20-04
sudo apt install python3-pip
sudo update-alternatives --install /usr/bin/pip pip /usr/bin/pip3 1
update pip (als normaler User):
pip install -U pip
OCR https://www.geeksforgeeks.org/license-plate-recognition-with-opencv-and-tesseract-ocr/?ref=lbp
pip install pytesseract
pip install opencv-python
sudo apt-get install tesseract-ocr tesseract-ocr-deu
cd ~/Software/OpenCV/detecting-and-ocring-digits
python ocr_digits.py --image apple_support.png
https://github.com/arturaugusto/display_ocr
Nice programm to change OCR parameters and see the effect immediately. (Slider for "Filter" not used in code ...)
Copied letsgodigital.traineddata
to /usr/share/tesseract-ocr/4.00/tessdata/
for global use.
Added some image manipulation functions. Needs fine tuned parameters to recognize my LCD photo! Screenshot:
https://wiki.ubuntuusers.de/gImageReader/
Screenshot: Working pretty well when cropping the right section manually :-)
See shell script test1/espcam_ocr.sh, reading all pic*.jpg from assets.
As well as the pictures seem to be pretty comparable, the results are heavily different...