-
Notifications
You must be signed in to change notification settings - Fork 0
/
region.py
309 lines (255 loc) · 9.89 KB
/
region.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
from abc import ABC, abstractmethod
import numpy as np
import os
os.environ["OPENCV_IO_MAX_IMAGE_PIXELS"] = pow(2, 40).__str__()
import cv2
class Flat(ABC):
def __init__(self, width: int, height: int):
self.bounds = tuple([width, height])
self.width, self.height = width, height
@property
def bounds(self) -> tuple[int, int]:
return self._width_height
@bounds.setter
def bounds(self, value: tuple[int, int]):
self._width_height = value
@property
def width(self) -> int:
return self._width
@width.setter
def width(self, value: int):
self._width = value
@property
def height(self) -> int:
return self._height
@height.setter
def height(self, value: int):
self._height = value
class LatLonSampler(ABC):
@abstractmethod
def sample(self, point: tuple[float, float]):
raise NotImplementedError
class Sampler(ABC):
@abstractmethod
def sample(self, point: tuple):
...
class FlatValues(Flat, Sampler, ABC):
def __init__(self, ground: np.array):
height, width = ground.shape[:2]
super().__init__(width, height)
self.values = ground
def set_pt(self, x, y, sample):
if min(self.height - 1, max(0, y)) != y or min(self.width - 1, max(0, x)) != x:
return
self.values[x, y] = sample
def sample(self, point: tuple[int, int]):
x, y = point
if min(self.height - 1, max(0, y)) != y or min(self.width - 1, max(0, x)) != x:
return 0, 0, 0
return self.values[min(self.height - 1, max(0, y)), min(self.width - 1, max(0, x))]
class EPSG32662(LatLonSampler): # WGS84 Plate Carrée
def __init__(self, fv: FlatValues, lat: tuple[float, float], lon: tuple[float, float]):
self.fv = fv
lat_from, lat_to = lat
lon_from, lon_to = lon
self.lat_flip = lat_from > lat_to
self.lon_flip = lon_from > lon_to
self.lat = np.linspace(lat_from, lat_to, num=fv.height, endpoint=True)
self.lon = np.linspace(lon_from, lon_to, num=fv.width, endpoint=True)
if self.lat_flip:
self.lat = np.flip(self.lat)
if self.lon_flip:
self.lon = np.flip(self.lon)
def sample(self, point: tuple[float, float]): # given in lat/lon
lat, lon = point
w = np.searchsorted(self.lon, lon) % self.fv.width
h = np.searchsorted(self.lat, lat) % self.fv.height
h = self.fv.height - h - 1 if self.lat_flip else h
w = self.fv.width - w - 1 if self.lon_flip else w
return self.fv.sample((w, h))
class CVFlatValues(FlatValues):
def __init__(self, samples: np.array):
super().__init__(samples)
@classmethod
def load(cls, f_name: str) -> 'CVFlatValues':
file = cv2.imread(f_name)
return cls(file)
@classmethod
def new(cls, dim: tuple[int, int]) -> 'CVFlatValues':
x, y = dim
blank = np.zeros((y, x), np.uint8)
space = cv2.cvtColor(blank, cv2.COLOR_GRAY2RGB)
return cls(space)
def poly(self, points, col):
c = np.array(col, dtype='uint8').tolist()
cv2.fillPoly(self.values, pts=points, color=c)
def show(self, wait: bool = True):
cv2.imshow('canvas', self.values)
if wait:
cv2.waitKey(0)
def save(self, fn: str = 'output'):
cv2.imwrite(f'output/{fn}.png', self.values)
class UnitSphere:
# Currently this is boundless - ie, all values can be reached.
# This can use various types of coordinate.
# I need to consider which is best.
# bearing in mind my primary function will be slerp,
# it seems to make sense to use unit vectors.
# self.xyz = tuple([vertices[d].xyz for d in self.vx]) # This is the unit sphere triangle.
def __init__(self, source: LatLonSampler):
self.source = source
@classmethod
def xyz_ll(cls, xyz: tuple) -> tuple[float, float]: # given a vector, return the latitude/longitude
x, y, z = xyz
return np.degrees(np.arctan2(z, np.sqrt(x * x + y * y))), np.degrees(np.arctan2(y, x))
def sample(self, point: tuple[float, float, float]): # given in spherical coordinates
ll = self.xyz_ll(point)
return self.source.sample(ll)
class Tri:
_tc = [
(3, 0, 2), (2, 0, 9), (2, 9, 1),
(6, 0, 5), (5, 0, 3), (5, 3, 4),
(9, 0, 8), (8, 0, 6), (8, 6, 7)
]
@classmethod
def slerp(cls, p0, p1, t):
# works only with unit vectors.
dot = np.dot(p0, p1)
th = np.arccos(dot)
s = np.sin(th)
j = np.sin((1. - t) * th)
k = np.sin(t * th)
return (j * np.array(p0) + k * np.array(p1)) / s
def __init__(self, ijk, abc):
# ijk is a tuple/list of three cartesian points on a unit sphere in clockwise order.
# abc is a tuple/list of three 2D cartesian points
# Need to calculate the seven remaining points, clockwise starting at the centre, then significant point (^ or V for up/down resp)
self.ijk = ijk # these are unit_sphere vectors.
self.abc = abc # These are equivalent 2D points.
self.c = np.mean(ijk, axis=0)
self.src = None
self.color = None
self.uvs = None
self.xyp = None
self.chn = []
def _set_uvs(self):
i, j, k = self.ijk
_pts = [
self.c, # centre.
i, # pt i
#
self.slerp(i, j, 1./3.),
self.slerp(i, j, 2./3.),
j, # p_j.
self.slerp(j, k, 1. / 3.),
self.slerp(j, k, 2. / 3.),
k, # p_k.
self.slerp(k, i, 1. / 3.),
self.slerp(k, i, 2. / 3.),
]
self.uvs = np.array(_pts) / np.linalg.norm(_pts, axis=1, keepdims=True) # each sub_triangle.
def _set_xyp(self):
i, j, k = self.abc
self.xyp = tuple([
np.mean([i, j, k], axis=0), # centre.
i, # pt i
np.mean([i, i, j], axis=0), # iij.
np.mean([i, j, j], axis=0), # ijj.
j, # p_j.
np.mean([j, j, k], axis=0), # jjk.
np.mean([j, k, k], axis=0), # jkk.
k, # p_k.
np.mean([k, k, i], axis=0), # kki.
np.mean([k, i, i], axis=0) # kii.
])
def create_chn(self, depth):
if depth <= 0 or self.uvs:
return
self._set_uvs()
self._set_xyp()
for a in self._tc:
uv = tuple([self.uvs[i] for i in a])
xy = tuple([self.xyp[i] for i in a])
c = Tri(uv, xy)
if self.src:
c.set_src(self.src)
c.create_chn(depth - 1)
self.chn.append(c)
def set_src(self, us: UnitSphere):
self.src = us
self.color = self.src.sample(self.c)
def poly_color(self):
result = []
if self.chn:
for c in self.chn:
result += c.poly_color()
else:
result.append(tuple([np.array([self.abc], dtype=np.int32), self.color]))
return result
def do_huge_spherical_triangle():
# This uses the central sub-triangle of the 'phi' map.
src = CVFlatValues.load(f'assets/90WE0_90N_21600.png')
globe = EPSG32662(src, (90., 0.), (-45., 45.))
width = 8000
height = int(np.round(np.sqrt(3.) / 2. * width))
disp = CVFlatValues.new((width+1, height+1))
# uk_uv = [[0.0, -0.5257311121191336, 0.85065080835204], [0.0, 0.5257311121191336, 0.85065080835204], [0.85065080835204, 0.0, 0.5257311121191336]]
uk_uv = [[0.63994974, -0.21203128, 0.73858451], [0.35682209, 0., 0.93417236], [0.63994974, 0.21203128, 0.73858451]]
uk_pt = [[0, height], [width >> 1, 0], [width, height]]
uk = Tri(uk_uv, uk_pt)
uk.set_src(UnitSphere(globe))
uk.create_chn(8)
pc_list = uk.poly_color()
for ppts, color in pc_list:
disp.poly(ppts, color)
disp.save('huge_triangle')
def do_big_flat():
src = CVFlatValues.load(f'assets/90WE0_90N_21600.png')
globe = EPSG32662(src, (90., 0.), (-45., 45.))
img_wid, img_hgt = 2400, 2400
fsr = CVFlatValues.new((img_wid, img_hgt))
la, lo = 51.58788025819696, -0.09624712666336593
dms = 15.0
lat_from, lat_to = la + dms, la - dms
lon_from, lon_to = lo - dms, lo + dms
lat = np.linspace(lat_from, lat_to, num=img_hgt, endpoint=True)
lon = np.linspace(lon_from, lon_to, num=img_wid, endpoint=True)
for x, lai in enumerate(lat):
for y, loi in enumerate(lon):
px = globe.sample((lai, loi))
fsr.set_pt(x, y, px)
fsr.show(True)
def do_flat_samples():
# import cv2
src = CVFlatValues.load(f'assets/world.topo.bathy.200406.3x5400x2700.png')
globe = EPSG32662(src, (90., -90.), (-180., 180.))
fsr = CVFlatValues.new((1000, 1000))
la, lo = 51.58788025819696, -0.09624712666336593
dms = 15.0
lat_from, lat_to = la + dms, la - dms
lon_from, lon_to = lo - dms, lo + dms
lat = np.linspace(lat_from, lat_to, num=1000, endpoint=True)
lon = np.linspace(lon_from, lon_to, num=1000, endpoint=True)
for x, lai in enumerate(lat):
for y, loi in enumerate(lon):
px = globe.sample((lai, loi))
fsr.set_pt(x, y, px)
cv2.imshow('fsr', fsr.values)
cv2.waitKey(0)
def do_spherical_triangle():
src = CVFlatValues.load(f'assets/world.topo.bathy.200406.3x5400x2700.png')
world = EPSG32662(src, (90., -90.), (-180., 180.))
width = 1000
height = int(np.round(np.sqrt(3.) / 2. * width))
disp = CVFlatValues.new((width+1, height+1))
uk_uv = [[0.0, -0.5257311121191336, 0.85065080835204], [0.0, 0.5257311121191336, 0.85065080835204], [0.85065080835204, 0.0, 0.5257311121191336]]
uk_pt = [[0, 0], [1000, 0], [500, height]]
uk = Tri(uk_uv, uk_pt)
uk.set_src(UnitSphere(world))
uk.create_chn(5)
pc_list = uk.poly_color()
for ppts, color in pc_list:
disp.poly(ppts, color)
disp.show(True)
if __name__ == '__main__':
do_spherical_triangle()