Using OpenCV and PyTesseract to localise text in scanned images such as invoice, reciepts, agreements
Create a folder in Root_Directory to store test images
Then, in command line execute the following commands, based on requirement
- python3 Text_Localisation-OpenCV.py
- python3 Text_Localisation-PyTesseract.py
PyTesseract performs better in scanned images when compared to opencv
When running Text_Localisation-OpenCV.py,change the element size in text_detect function to get a different sized bounding boxes. Smaller ele_size corresponds to more accurate detection(like letters)