I have a questions regarding the selection of the outer points (i.e. sphere center), among a collection of many spheres in 2D/3D space.
The outer points is that when all of these points are connected in a closed envelope, the rest of points are inside the circle.
Assuming the spheres center (x, y, z) and radius are given and stored in an array named spheresN:
x1, y1, z1, rad1;
x2, y2, z2, rad2;
...
xN, yN, zN, radN;
I tried to make a bigger box that contains all spheres inside it, and calculate the X-distance/Y-distance to the box walls for all spheres. But could not figure out what criteria should be used to judge the sphere is or is NOT the outer.
Any advice or recommendation is highly appreciated.
Thanks in advance!
Update 1:
I did try some of the convex Hull alorithms, but the resulting points are ONLY the minimum points for convex envelope containing all of the points inside. Below is a example I tested: 10 points, their x, y coordinates are between [0, 1]. The envelope does NOT contain all outer points, it's just the envelope with minimum outer points!
So literally what I need is to find the envelope/circle with maximum points.
Update 2:
The motivation of this question is: I have done some computer simulation regarding real particle flow. Below is a model of real particle using overlapping spheres (via scanned 3D surface mesh). I want to select the spheres on the surface of this particle as a group.
So the sphere center (x,y,z) and sphere radius have to be taken into account in the selecting processing. However the convex hull algorithm only calculates points (radius = 0). So I think the convex hull algorithm is not working in this case any more.
Any ideas?
Update 3:
As I did some research on the Hull finding algorithms out of a set of points, I think what I am looking for is the concave Hull algorithm in 2D and 3D.
e.g. a set of points - shape "H"
