3 ms·
When I experimented with 2D metaballs a decade ago I started with a low resolution scan and then subdivided all squares whose corners weren't either all inside
by ema 6y ago
When I experimented with 2D metaballs a decade ago I started with a low resolution scan and then subdivided all squares whose corners weren't either all inside or outside until I had individual pixels. Except for a few glitches when just checking the corners wasn't enough this worked rather well and was easy to implement.