Click space to scan the current camera view, and then it'll draw it out. The more details in the camera view, the longer the scan will take. Still in testing phase, so check out the code to see how it works. (I'll be adding comments soon). The sub-two-minute times I've been getting are primarily thanks to an algorithm I added to the color detection script. This algorithm creates a color cache, so colors used multiple times will be found much faster.
Thanks to @frodewin for the color detection script. Check out that project here: https://scratch.mit.edu/projects/13412067/ Everything else was created by me.
