The tesseract is an OCR engine (model) from Google.
The supports multi languages, with multiple models
The performance is not too great comparing to a self trained EAST + CRNN solution.
The tesseract is trainable, so maybe given more data it will perform better.
#install tesseract
#there is windows installer
#add the install folder to path environment variable
#the install files include the model file
#pytesseract is only a wrapper
from PIL import Image
import pytesseract
import cv2
file_name = r'C:\data\test data\bunnings.jpg'
# img = Image.open(file_name)
img = cv2.imread(file_name)
res = pytesseract.image_to_string(img)
res = res.replace('\x0c', '')
print(res)