jet@karazm.math.uh.edu (J. Eric Townsend) (09/17/90)
Well, maybe discrete isn't the right word, but it sounds good... :-) I have objects built from triangle lists (currently unsorted, could spatially sort if I cared). Each triangle is described by it's vertexes. (Vertexi?) The question is, how to quickly build octrees around these. I'm currently writing my code to do the following: 1. Determine if triangle intersects with cube. Do six plane-plane tests, if any intersections exist check for intersection point to be within a wall of the box. 2. Check for abitrary point on triangle to be within box. If 2, but not 1, then the triangle is completely within the box. If 1 and 2, then the triangle intersects with the box. This will be a terminal node. I've not figured out a way to determine if the box is within the triangle, so I stop subdividing when I hit an intersection of box and triangle. Is there a quicker way to do this? I'm interested in speed because I suspect (but don't know for sure) that it's quicker to manipulate the triangles (movement, shape munging, etc) than to screw around with transforming big octrees. (If I'm wrong, feel free to let me know. :-) I plan on comparing the speeds of (build octree + trace) and (build spherical bounding volumes and trace). Thx for any help. [Obligatory non-seq: Hey Glassner, nice book. Worth the $$$ for the biblio alone.] -- J. Eric Townsend -- University of Houston Dept. of Mathematics (713) 749-2120 Internet: jet@uh.edu Bitnet: jet@UHOU Skate UNIX(r)
jet@karazm.math.uh.edu (J. Eric Townsend) (09/18/90)
Just to clarify my earlier request. In article <1990Sep17.073033.10458@lavaca.uh.edu> jet@karazm.math.uh.edu (J. Eric Townsend) writes: >I have objects built from triangle lists (currently unsorted, >could spatially sort if I cared). Each triangle is described by >it's vertexes. (Vertexi?) I'm dealing with objects built of triangles (for now). I'm not very particular about keeping track of the object's place within the octree -- I'm more interested in where the pieces of the object are. For the person who sent me mail (I lost it): I'm not illiterate just because I don't know the plural of a math-word. -- J. Eric Townsend -- University of Houston Dept. of Mathematics (713) 749-2120 Internet: jet@uh.edu Bitnet: jet@UHOU Skate UNIX(r)