I created this project mainly for fun. I saw some examples of pixel sorting online and thought that it would be pretty fun and easy to replicate it! The project was created in python, using primarily numpy and pillow to open and process through images.
It works by opening images and converting them into numpy arrays. The user specifies a pixel attribute to evaluate (ex: pixel intensity, hue, or saturation), as well as upper and lower thresholds for that attribute. The program then iterates through each row of the the array and denotes intervals of pixels that are within those thresholds. Then, for each interval, the pixels are sorted, usually by the same attribute used for determining intervals. The result is really cool images that look like they’re melting away. Below is a gallery of my favorite image created by the script!
 |
 |
 |
 |
 |
 |
 |
 |
 |