I recently learnt about SVD (Singular Value Decomposition) and the example application our lecturer gave was image compression.
I thought this was pretty neat so I made my own in python.
(I was so close to doing it in C, but I couldn't be bothered)
I got one of my mandelbrot images and compressed that, so sit back and enjoy the soothing sights of highly compressed mandelbrot.
Original:
One singular value: (Highest compression)
Here's a few more, increasing by 1 value each time
After having fun with compression, I decided to play around with changing the singular values a little and found some interesting results
I made the following 2 by only taking every second value of the decomposition:
This is my favourite :)