In this video, I tried to show how patches (small portions) can be created from a big image in two different ways: sequentially and randomly. For creating patches sequentially, I used the 'patchify' library; for the random case, I used the 'NumPy' library. We don't get any overlapping when we extract patches sequentially but overlaps are expected in case of random extraction.
GitHub address: github.com/ran...
#datascience #python #jupyternotebook #numpy #matplotlib #patchify #patchcreation
17 окт 2024