Thanks I’ve got no idea how to do that in ICE. I’ll probably go with scripting even if it will be much slower.
Martin Sent from my iPhone > On Oct 25, 2017, at 17:46, Matt Lind <[email protected]> wrote: > > Comparing the UV coordinates at each vertex is generally correct, but it > sounds like the comparison wasn't deep enough. > > Consider this: > > for each edge of mesh: > If edge == boundary: > UV coordinates of edge are UV island border > else: > if edge's UV coordinates at vertex 0 are same value AND > edge's UV coordinates at vertex 1 are same value AND > sign of polygon A's determinant == sign of polygon B's determinant: > edge is an interior edge of same UV Island. > > Use UV coordinate values of the first 3 vertices comprising the polygon, in > proper winding order, to compute the determinant. > > Overall level of difficulty is low, but Softimage data structures are not > ideal for this task as they're designed for efficient drilling downwards > (e.g. object -> polygon -> edge -> vertex -> polygon node). There's little > support for going the other direction (e.g. polygon node -> vertex -> > edge -> polygon -> object). You'll spend more time figuring out how to > traverse the mesh efficiently than on the problem of which UV island a > polygon node belongs. There are multiple solutions, but running time varies > significantly. > > Matt > > > Date: Wed, 25 Oct 2017 16:56:19 +0900 > From: Martin Yara <[email protected]> > Subject: UV Island Border > To: "[email protected]" > > Any idea how to get the island border through scripting ? (or ICE) > > I was using a compound rray uploaded in si-community where the logic is > comparing every edge's sample UV positions. If they don't share the same > position, or is a border edge, then it's a UV border. > > The problem with this is that it doesn't recognize the borders of islands > that are symmetric and overlapped (like folded in the middle). > > Any ideas ? > > I'm trying to reproduce a tool I wrote in Maya to verify the distance > between all the islands to check if they are not too close and cause > bleeding problems. > > Martin > > > ------ > Softimage Mailing List. > To unsubscribe, send a mail to [email protected] with > "unsubscribe" in the subject, and reply to confirm. ------ Softimage Mailing List. To unsubscribe, send a mail to [email protected] with "unsubscribe" in the subject, and reply to confirm.
