| "Rules: Logic and Applications" 2nd Workshop, Dec,
2019 |
| Aesthetic Morphisms |
| Jocelyn Ireson-Paine |
| www.jocelyns-cartoons.uk/rules2019/ |
Cartoon Enhancement Results
Tested on several figures similar to
those in my training set, the learned
model seemed to generate reasonable results.
BUT ... inspection showed that
the reds were impure, so that channel
separation left some smears of black
behind.
My
model may
just have learnt to convert these
back to red! Because of my slow
machine, I've not investigated this.
I tried again, this time drawing
training images in Gimp with pure
red and pure black. I'm not good
at drawing with a mouse, so the pictures
were crude. But black-red separation
was complete, and I still got reasonable
results.
