- Clone the repo on your computer
- Run
pip install imutils
anywhere - Run
pip install numpy
anywhere - Run
pip install opencv-python
anywhere - Run
python opencv_text_detection_image.py --image images/Book.jpg --east frozen_east_text_detection.pb
in the local repo
- Put the input images in the
input
folder - The output images will be in the
output
folder in sorted order
- Increase resized width and height for better accuracy (current default:
1280
) - Cons: Takes more time the more you increase the size. At
1280
it takes almost 5 seconds - Pros: Increases accuracy a lot. At
1280
accuracy is very high alhamdulillah