Math + Making

A student blog for Math 189AH: Making Mathematics at Harvey Mudd College

Digital Photos and Kernel Convolutions

Ambika Tiwari

In this project, I explored how images are stored in computers as arrays, and how we use math to manipulate those images using kernel convolutions to create photography filters, optimize machine learning models, and get a beautiful visual intuition of math with matrices. I used a Google Colab notebook to go step-by-step to get used to seeing images as pixels, to see exactly what the math is and how it can be represented with code, and I’ve commented and explained along the way. For the “making” aspect, I wanted to make something that was interactive, so I hope that you clone the notebook – tinker with the convolution function, upload your own images, and try out different kernels!

Please find the rest of my blog and what I made in the notebook!


Leave a Reply

Your email address will not be published. Required fields are marked *