如果凸包是完美的矩形(python 3),如何得到构成凸包的所有点?

时间:2018-06-16 13:58:23

标签: python convex-hull convex-polygon qhull

打开图片查看以下代码的结果

import numpy as np
from scipy.spatial import ConvexHull
import matplotlib.pyplot as plt

points = np.array([[1,1],[1,2],[1,3],[1,4],[2,1],[2,2],[2,3],[2,4],[3,1],[3,2],[3,3],[3,4],[4,1],[4,2],[4,3],[4,4]])  
hull = ConvexHull(points)
plt.plot(points[:,0], points[:,1], 'o')
for simplex in hull.simplices:
    plt.plot(points[simplex, 0], points[simplex, 1], 'k-')
    plt.plot(points[simplex,0], points[simplex,1], 'ro', alpha=.25, markersize=20)

我想获得凸包上点的坐标索引(黑点+在线的点)。我选择矩形只是为了得到极端情况。

hull.points 只能提供标记为红色的点(仅限矩形的角点)。

result of code

1 个答案:

答案 0 :(得分:0)

如果您确定凸包是一个完美的矩形,其边与x轴和y轴对齐,找到所有边界点的索引很简单。凸壳完全不需要计算。那个描述适合你的例子。以下是一些代码,可以在这种情况下执行您想要的操作。此代码的时间复杂度为O(n),其中n是整体的点数。

# Find the indices of all boundary points, in increasing index order,
#   assuming the hull is a rectangle aligned with the axes.
x_limits = (min(pt[0] for pt in points), max(pt[0] for pt in points))
y_limits = (min(pt[1] for pt in points), max(pt[1] for pt in points))
boundary_indices = [idx for idx, (x, y) in enumerate(points) 
                    if x in x_limits or y in y_limits]

然而,这种情况似乎很简单。以下是适用于所有二维情况的更通用的代码,尤其是当这些点具有整体坐标时。这是因为如果精度不准确,找出一个点是否正好在一个线段上是很棘手的。此代码以时间复杂度O(n*m)运行,其中n是点数,m是凸包中的顶点数。

# Find the indices of all boundary points, in increasing index order,
#   making no assumptions on the hull.
def are_collinear2d(pt1, pt2, pt3):
    """Return if three 2-dimensional points are collinear, assuming 
    perfect precision"""
    return ((pt2[0] - pt1[0]) * (pt3[1] - pt1[1]) 
          - (pt2[1] - pt1[1]) * (pt3[0] - pt1[0])) == 0

vertex_pairs = list(zip(vertices, vertices[1:] + vertices[0:1]))
boundary_indices = []
for idx, pt in enumerate(points):
    for v1, v2 in vertex_pairs:
        if are_collinear2d(pt, v1, v2):
            boundary_indices.append(idx)
            break