Created with Python 3.11.3
pip install -r requirements.txtThe original code is from https://github.com/techwithtim/PythonPaintProgram with some small modifications made to fit the MNIST dataset.
The Visualtier takes in the specified file as an image and colvolves with the selected kernel. The output is normalized to a range between 0-255 to fit a graysclaed image and also converted to red and green to show the detection with a better contrast.
Press "ESC" or the X in the top right corner to close the window.
See settings.py for the available kernels or create your own.