Honestly, I’ve been wanting a block like this for so long and I’m so happy its finally here. Feel like this would be a great fit for library.
when i run this script
the costume that is added to the costume list in not and svg
this would be much cooler if it was all written in snap! code 
It also would be near impossible to run. I’ve designed it kinda. It would require smoothing image (blur), k mean shift Clustering, then for every color locate each individual shape (can be done by detecting neighbors for each pixel and grouping), then recognizing straight lines and curves (thresheld), calculating the total outline of the shape, apply the recognition of straight and curved lines to total outline, balloon each shape (math), slice each shape by outline shape to prevent overflow, draw a image with the largest shapes ballooned a lot and with each shape size make the ballooning smaller, then draw the regular image over for accuracy.
I’m pretty sure they wanted a bitmap in an svg format.
bitmap in svg format is just bitmap…