如何拼合选择的点数

时间:2013-03-06 13:12:35

标签: 3d maya

我需要在其LOCAL法线轴上展平选择的点。我假设如果法线是正确的,那么无论它们是从对象的任何一侧选择点,这个轴总是相同的吗?

为了直观地表达我想要实现的目标,我想转而这样:

enter image description here

以编程方式进入:

enter image description here

如果我将比例工具设置为'法线平均值',并手动缩放它们,我可以将点展平为平面,但我需要通过代码计算或执行此操作。

我查看了polyMoveFacet命令,它有一个名为localScaleZ的标志,甚至在其描述中写有“Flattening”,但我没有运气。有什么建议吗?

2 个答案:

答案 0 :(得分:2)

最简单的方法就是手动使用同样的东西。在mel中执行此操作的代码如下所示:

{ // protect global namespace
     setToolTo Scale;
     manipScaleContext -e -mode 9 Scale;
     $oa = `manipScaleContext -q  -orientAxes Scale`;
     $p = `manipScaleContext -q  -position Scale`;
     scale -ws -r 
           -p ($p[0]) ($p[1]) ($p[2]) 
           -oa ($oa[0]+"rad") ($oa[1]+"rad") ($oa[2]+"rad") 
            0 1 1;
}

和Python:

cmds.setToolTo('Scale')
cmds.manipScaleContext("Scale", e=1, mode=9)
p = cmds.manipScaleContext("Scale", q=1, p=1)
oa = cmds.manipScaleContext("Scale", q=1, oa=1) 
cmds.scale(0,1,1, 
           p=(p[0],p[1],p[2]),
           oa=("%srad"%oa[0],"%srad"%oa[1],"%srad"%oa[2]))

答案 1 :(得分:2)

我从未使用过maya,也不知道它使用的脚本语言。因此,这个答案只涉及问题的数学/几何方法。代码在python中用于演示概念,但您应该能够翻译。

请注意,我没有测试代码,但希望它至少可以为您提供解决问题的工具。

from math import sqrt

def point_average(points):
    l = len(points)
    return [(p.x/l,p.y/l,p.z/l) for p in points]

def find_normal(points):
    normal = point_average([p.normal for p in points])
    normal_length = sqrt(sum(c**2 for c in normal))
    normal = [c/normal_length for c in normal]
    return normal

def find_plane(points):
    normal = find_average_normal(points)
    center = point_average(points)
    # a point and a normal are enough to uniquely identify a plane
    # we anchor the plane to the farthest point from the center
    # that should be one of the corners
    dcenter = lambda p:sqrt((p.x-center.x)**2+(p.y-center.y)**2+(p.z-center.z)**2)
    points = [(dcenter(p),p) for p in points]
    points.sort()
    anchor = points[-1][1]
    return (anchor,normal)

def project_point_onto_plane(point, plane):
    anchor,normal = plane
    # kudos to http://stackoverflow.com/questions/9605556/how-to-project-a-3d-point-to-a-3d-plane
    # for the math behind this
    v = (point.x-anchor[0], point.y-anchor[1], point.z-anchor[2])
    dist = v[0]*normal[0] + v[1]*normal[1] + v[2]*normal[2]
    projected_point = (point.x-dist*normal[0],
                       point.y-dist*normal[1],
                       point.z-dist*normal[1])
    return projected_point