2D Fourier Transform Drawing Machine

Motivation

While browsing 3Blue1Brown — a math YouTube channel I follow closely — I came across a video explaining how to approximate any arbitrary path using the Fourier Transform on the complex plane. The idea of decomposing a shape into a sum of rotating circles was immediately compelling. My goal for the AE280 final project was to reproduce this from scratch in Python.

How It Works

Rather than taking vector paths as input, the implementation processes ordinary raster images (JPG/PNG). Edge detection extracts the contour, which is then sampled as a sequence of complex numbers. Applying the DFT yields a set of frequency components; animating these as rotating phasors reconstructs the original shape as a sum of circles.

Results

GitHub