[Soc-2016-dev] Weekly Report #7, UV-Tools

[Phil Gosch]

Report #7 for UV-Tools:

What I did this week:

* Worked on the implementation of a single packing solution. This involved the computation of the Inner Fit Polygon per UV chart, which basically is a modified No Fit Polygon which makes sure no items are placed outside the UV bounds. I was able to simplify this compuation since in our case the UV bounds are always square/rectangular.

I then implemented the packing algorithm using NFPs/IFPs by choosing random placement positions for testing purposes. My mentor adviced me to hold of with NFP union computation (for getting a correct Collision Free Region) and just use simple testing with random choosen points on the NFPs, to first get the rest of the algorithm working. Right now there's a bug with NFP placement, due to finding the correct reference vertices which I expect to solve soon (maybe even over the weekend). Finding the right reference vertices for chart placement turned out to be way more important than I first thought btw.

I also did a lot of refactoring and cleanup of the current implementation, although it still contains lots of debug prints, which I'll only remove once everything is working.

* Implemented a "Select Overlapping UVs" operator since it was requested by a lot of users. [1]

What I plan on doing next week:

* Finish the last bits of single packing solution computation, test and clean up the code.

* Once that's done I'll start working on the simulated annealing metaheuristic. I expect this to not take too long, however the hard part is going to be finding "good" parameters, which will take lots of testing and balancing.

Questions:

I'm computing the "% used UV space" during packing (of course in it's own function). How can I store this variable and permanently display to the user (for example via toolbar)? (already got requests for that by users)

[1] http://www.pasteall.org/pic/show.php?id=104792

-- pixel-pusher at saphirestudio.at 
_______________________________________________ Soc-2016-dev mailing 
list Soc-2016-dev at blender.org 
https://lists.blender.org/mailman/listinfo/soc-2016-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/soc-2016-dev/attachments/20160708/d04ff79e/attachment.htm