我正在整形地使用以找到两个椭圆的重叠区域。
这是椭圆函数：
import torch
from shapely.geometry.point import Point
from shapely import affinity
def create_ellipse(center, lengths, angle=0):
"""
create a shapely ellipse. adapted from
https://gis.stackexchange.com/a/243462
"""
circ = Point(center).buffer(1)
ell = affinity.scale(circ, int(lengths[0]), int(lengths[1]))
ellr = affinity.rotate(ell, angle)
return ellr
有2个pytorch张量：
wh1 = torch.tensor([3.6250, 2.8125])
w1 = wh1[0]
h1 = wh1[1]
wh2 = torch.tensor([[ 1.9215, 8.2504, 0.2698, 2.1242, 0.3143, 2.1111, 1.9065, 1.3590,
1.9489, 1.3957, 0.7099, 1.2371, 0.9294, 0.3765, 0.3079, 0.2556,
0.2194, 0.2841, 0.3109, 0.8383, 0.3096, 0.2565, 0.3197, 0.5158,
0.3186, 0.3041, 0.5445, 0.2659, 0.5199, 7.7250, 6.2023, 7.0652,
0.6808, 0.8072, 1.4054, 0.7802, 2.3664, 2.8730, 0.4354, 1.4630,
4.7688, 4.6711, 1.5979, 2.4067, 1.7955, 1.8868, 1.8641, 1.5332,
1.5486],
[ 1.7919, 4.1976, 0.3816, 1.9401, 0.4061, 2.7557, 2.5397, 1.7599,
1.8994, 1.5508, 0.4886, 1.1070, 0.9757, 0.7873, 0.4764, 0.3846,
0.4159, 0.4954, 0.3313, 0.8940, 0.5923, 0.3516, 0.5340, 0.3600,
0.4161, 0.2797, 0.8646, 0.5811, 0.4164, 0.8625, 5.2044, 10.5977,
0.8569, 0.6457, 2.4552, 0.8479, 3.8021, 2.6354, 1.2007, 1.7143,
3.6156, 7.8706, 2.1638, 7.8473, 1.9153, 2.0610, 2.2354, 1.9746,
1.9436]])
w2 = wh2[0]
h2 = wh2[1]
w1和w2是椭圆的宽度，h1和h2是椭圆的高度。我想找到第一个椭圆（w1和h1）以及每个带有w2和h2的椭圆的重叠区域。输出的大小应与w2或h2相同。我运行以下命令：
ell1 = create_ellipse((0, 0), (w1.item() / 2.0, h1.item() / 2.0), 0)
for w2_t, h2_t in zip(w2, h2):
print('w2: ', w2_t.item(), 'h2: ', h2_t.item())
ell2 = create_ellipse((0, 0), (w2_t.item() / 2.0, h2_t.item() / 2.0), 0)
inter_area = ell1.intersection(ell2).area
print('inter area: ', inter_area)
然后显示以下错误：
TopologyException: Input geom 1 is invalid: Self-intersection at or near point -1 0 at -1 0
w2: 1.9214999675750732 h2: 1.7919000387191772
inter area: 0.0
w2: 8.250399589538574 h2: 4.1975998878479
inter area: 3.1365484905459384
w2: 0.26980000734329224 h2: 0.3815999925136566
inter area: 0.0
w2: 2.1242001056671143 h2: 1.9400999546051025
---------------------------------------------------------------------------
TopologicalError Traceback (most recent call last)
<ipython-input-143-68e4119a1c74> in <module>()
3 print('w2: ', w2_t.item(), 'h2: ', h2_t.item())
4 ell2 = create_ellipse((0, 0), (w2_t.item() / 2.0, h2_t.item() / 2.0), 0)
----> 5 inter_area = ell1.intersection(ell2).area
6 print('inter area: ', inter_area)
2 frames
/usr/local/lib/python3.6/dist-packages/shapely/geometry/base.py in intersection(self, other)
647 def intersection(self, other):
648 """Returns the intersection of the geometries"""
--> 649 return geom_factory(self.impl['intersection'](self, other))
650
651 def symmetric_difference(self, other):
/usr/local/lib/python3.6/dist-packages/shapely/topology.py in __call__(self, this, other, *args)
68 err = TopologicalError(
69 "This operation could not be performed. Reason: unknown")
---> 70 self._check_topology(err, this, other)
71 return product
72
/usr/local/lib/python3.6/dist-packages/shapely/topology.py in _check_topology(self, err, *geoms)
36 "The operation '%s' could not be performed. "
37 "Likely cause is invalidity of the geometry %s" % (
---> 38 self.fn.__name__, repr(geom)))
39 raise err
40
TopologicalError: The operation 'GEOSIntersection_r' could not be performed. Likely cause is invalidity of the geometry <shapely.geometry.polygon.Polygon object at 0x7f7bc3100940>
我发现了问题。函数：
def create_ellipse(center, lengths, angle=0):
"""
create a shapely ellipse. adapted from
https://gis.stackexchange.com/a/243462
"""
circ = Point(center).buffer(1)
ell = affinity.scale(circ, int(lengths[0]), int(lengths[1]))
ellr = affinity.rotate(ell, angle)
return ellr
应更改为：
def create_ellipse(center, lengths, angle=0):
"""
create a shapely ellipse. adapted from
https://gis.stackexchange.com/a/243462
"""
circ = Point(center).buffer(1)
ell = affinity.scale(circ, lengths[0], lengths[1])
ellr = affinity.rotate(ell, angle)
return ellr