10楼: Originally posted by chujian77 at 2012-05-15 22:41:15:
我的看法是：
1.给出包含T型区域的三个虚拟点（坐标够大即可），加入到T型区域点集中，进行剖分。
2.找出T型边界与你的剖分网的交点，作为新点加入到原来的点集，重新再做一次Delaunay剖分，基本上可以保证T型区 ...

11楼: Originally posted by acmuser at 2012-05-15 23:05:09:
that may not work, run this matlab script, n is the number of points in the cloud, and R parametrize the virtual points,
function delaunayclouds(n,R)

x=rand(n,1);
y=rand(n,1);

ir = find(x ...

12楼: Originally posted by acmuser at 2012-05-15 23:14:49:
I can even add more virtual points, but still no help, in the following, I add 8 virtual points
16/f5/1582407_1337094886_789.jpg

13楼: Originally posted by acmuser at 2012-05-15 23:22:15:
I think there are some boundary estimate or support estimate algorithm for data clouds in computer graphics, that might help to properly solve this problem. For example,
Rolling-Ball Method For E ...