Division of numbers from contours in opencv in python for cnn

I want to separate numbers in suppose 7638 into different images which can be predicted individually using cnn.
By finding contours how can I divide each contour into separate image in python.
To be more clear: How can I divide https://www.letter2word.com/products/americana-numbers into 0,1,2,3..9 into individual images using the concept of contours in opencv in python so that the resulting images can be predicted by cnn classifier and later all the results written together to give the number in image(initial) as output.
Also, is there any other method to do so.
PS.-I tried searching in google but everywhere it is given to extract 1 particular part of the image while I need more than 1 parts of the image for correct prediction