Real-Time Physics Simulation Forum

Hi, If I understand you correctly, you're encountering the same situation I am with the special case of the end points of the arc. The enclosing 'tile' test is really trivial, its the OBB vs arc that gets tricky, since the endpoints of the arc present a special case. To your comment about the differ...

Interesting, Im not familiar with that technique. I googled around a bit and could only find paper abstracts with regards to collision. christer ericson refers a bit to it in his book, but not in depth. Does it provide a projection depth/vector? I would be happy to just have an object vs arc segment...

Hi Raigan, I posted an image to show what the problem case is. http://aycu30.webshots.com/image/7269/2001872211888135380_th.jpg There's clearly a collision on all three "axes" but the rect vertex inside the voronoi region is not at a distance greater than the arc radius. In using an AABB t...

Thanks for the reply. I actually cross posted to your forums so feel free to ignore that one. With regards to the example on your site, yeah, I did notice that the arc penetration gets dumped if the relevant box vertex is outside the bounds of the rect enclosing the arcsegment. There are certain sit...

Im trying to implement the SAT between an OBB and a concave 90 degree arc segment, in 2D. To understand what this looks like, check out the tutorial on the SAT here: http://www.harveycartel.org/metanet/tutorials/tutorialA.html (figure 7) Figure 7 shows the technique to check for a collision between ...