I have personally collected this Dataset.
'ImageSegmentation.py' is code to perform segmentation -> Separate individual digit from image containing multiple digits. 'JPG_to_CSV.py' is a code to convert greyscale images into .csv file.
'HWDR- Traninig.py'-> It takes training dataset, and store trained parameters in .npz file. 'HWDR- Perform Recognition'-> It uses .npz file, generated in earlier step. and then perform recognition on provided data