[Tutorial] Recognize Handwritten Numbers in Python without any ML-Library

Written by joelbarmettlerUZH | Published 2019/10/15
Tech Story Tags: learn-python | python-basics | learn-programming | learn-artificial-intelligence | learn-machine-learning | python-programming | latest-tech-stories | hackernoon-top-story

TLDR The image classifier is a simple python script that recognizes a hand-written number. The input image shall all have a fixed size of 64x64 pixels, the images are RGB coloured. The image-category with the highest similarity will serve us as an estimate. When a new, unknown image is provided, the algorithm loops through the database and computes the similarity to the well-known images. We will NOT use any machine learning library like Tensorflow, Keras or SkLearn. To open images, we will use the Image class from the PIL (Python Image Library)via the TL;DR App

no story

Written by joelbarmettlerUZH | Founder of coinpaper.io - Crypto Info, Price, Review and Analysis
Published by HackerNoon on 2019/10/15