this is ugly, experimental, slow, unoptimized code. its purpose is to verify my (propably) stupid idea of an "image to hatching pen plotter drawing". The usual warning applies: I have no idea what I am doing.
for now the script finds edges and fills areas in between with colors
edge detection works as expected, but drawn edges (black lines...) are not handled in a special way = handled like a normal area/fill = results in double lines detected for comic like input drawings
motion vector recovery works ok-ish.
this script now generates lots of data from an image file
next, yet unimplemented, step is to combine all this data and generate polygon lines
another, yet unimplemented, part is to write all the polygons out in HPGL form.
cleanup and split-up in multiple files in in progress (seperate library/supporting stuff from high level logic)