Code tested on : Python version 3.6.1 opencv-python 3.4.7+contrib tesserocr 2.4.0 (tesseract v4)
->Get tesserocr from here : https://github.com/simonflueckiger/tesserocr-windows_build/releases From the download directory, open a command prompt (simply point it to the directory that holds the ".whl" file if you opened a command prompt from other directory). Installation via pip is done via the following code:
pip install <package_name>.whl
Package_name refers to the name of the whl file you have downloaded. In my case, I have downloaded tesserocr-2.4.0-cp36-cp36m-win_amd64.whl. Hence, I will be using the following code for the installation:
pip install tesserocr-2.4.0-cp36-cp36m-win_amd64.whl
->Get tesseract's language data files from here : https://github.com/tesseract-ocr/tessdata For English, it is "eng.traineddata"
->Place "eng.traineddata" inside tessdata folder. The folder should be like:
./tessdata :
-configs
-tessconfigs
-eng.traineddata
-eng.user-patterns
-eng.user-words
-info.txt
-osd.traineddata
-pdf.ttf
->Put the input images in "LicPlateImages" folder.
->Run Main_v1_image final.py Output images go in the "output" folder, thresholded images and output text is saved in the "thresimages" folder.