/
thresholding.py
262 lines (206 loc) · 10.2 KB
/
thresholding.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
import ee
from ee.ee_exception import EEException
import random
from hydrafloods import geeutils, decorators
def bmaxOtsu(collection,
band=None,
reductionScale=90,
initialThreshold=0,
invert=False,
gridSize=0.1,
bmaxThresh=0.75,
maxBoxes=100,
seed=7):
@decorators.carryMetadata
def applyBmax(img):
def constuctGrid(i):
def contructXGrid(j):
j = ee.Number(j)
box = ee.Feature(ee.Geometry.Rectangle(j,i,j.add(gridSize),i.add(gridSize)))
out = ee.Algorithms.If(geom.contains(box.geometry(),maxError=1),box,None)
return ee.Feature(out)
i = ee.Number(i)
out = ee.List.sequence(west,east.subtract(gridSize),gridSize).map(contructXGrid)
return out
def calcBmax(feature):
segment = img#.clip(feature)
initial = segment.lt(initialThreshold)
p1 = ee.Number(initial.reduceRegion(
reducer= ee.Reducer.mean(),
geometry= feature.geometry(),
bestEffort= True,
scale= reductionScale,
).get(histBand))
p1 = ee.Number(ee.Algorithms.If(p1,p1,0.99))
p2 = ee.Number(1).subtract(p1)
m = segment.updateMask(initial).rename('m1').addBands(
segment.updateMask(initial.Not()).rename('m2')
)
mReduced = m.reduceRegion(
reducer= ee.Reducer.mean(),
geometry= feature.geometry(),
bestEffort= True,
scale= reductionScale,
)
m1 = ee.Number(mReduced.get('m1'))
m2 = ee.Number(mReduced.get('m2'))
m1 = ee.Number(ee.Algorithms.If(m1,m1,-25))
m2 = ee.Number(ee.Algorithms.If(m2,m2,0))
sigmab = p1.multiply(p2.multiply(m1.subtract(m2).pow(2)))
sigmat = ee.Number(segment.reduceRegion(
reducer= ee.Reducer.variance(),
geometry= feature.geometry(),
bestEffort= True,
scale= reductionScale,
).get(histBand))
sigmat = ee.Number(ee.Algorithms.If(sigmat,sigmat,2))
bmax = sigmab.divide(sigmat)
return feature.set({'bmax':bmax})
geom = img.geometry()
bounds = geom.bounds(maxError=1)
coords = ee.List(bounds.coordinates().get(0))
gridRes = ee.Number(gridSize)
west = ee.Number(ee.List(coords.get(0)).get(0))
south = ee.Number(ee.List(coords.get(0)).get(1))
east = ee.Number(ee.List(coords.get(2)).get(0))
north = ee.Number(ee.List(coords.get(2)).get(1))
west = west.subtract(west.mod(gridRes))
south = south.subtract(south.mod(gridRes))
east = east.add(gridRes.subtract(east.mod(gridRes)))
north = north.add(gridRes.subtract(north.mod(gridRes)))
grid = ee.FeatureCollection(
ee.List.sequence(south,north.subtract(gridRes),gridRes).map(constuctGrid).flatten()
)
bmaxes = grid.map(calcBmax).filter(ee.Filter.gt('bmax',bmaxThresh)).randomColumn('random',seed)
nBoxes = ee.Number(bmaxes.size())
randomThresh = ee.Number(maxBoxes).divide(nBoxes)
selection = bmaxes.filter(ee.Filter.lt('random',randomThresh))
histogram = img.reduceRegion(ee.Reducer.histogram(255, 1)\
.combine('mean', None, True)\
.combine('variance', None,True),selection,reductionScale,bestEffort=True,
tileScale=16)
threshold = otsu(histogram.get(histBand.cat('_histogram')))
water = ee.Image(ee.Algorithms.If(invert,img.gt(threshold),img.lt(threshold)))
return water.rename('water').uint8()
if band is None:
collection = collection.select([0])
histBand = ee.String(ee.Image(collection.first()).bandNames().get(0))
else:
histBand = ee.String(band)
collection = collection.select(histBand)
return collection.map(applyBmax)
def edgeOtsu(collection,
initialThreshold=0,
canny_threshold=0.05, # threshold for canny edge detection
canny_sigma=0, # sigma value for gaussian filter
canny_lt=0.05, # lower threshold for canny detection
smoothing=100, # amount of smoothing in meters
connected_pixels=200, # maximum size of the neighborhood in pixels
edge_length=50, # minimum length of edges from canny detection
smooth_edges=100,
band=None,
reductionScale=90,
initThresh=0,
invert=False,
seed=7):
@decorators.carryMetadata
def applyEdge(img):
# get preliminary water
binary = img.lt(initialThreshold).rename('binary');
# get canny edges
canny = ee.Algorithms.CannyEdgeDetector(binary,canny_threshold,canny_sigma)
# process canny edges
connected = canny.mask(canny).lt(canny_lt).connectedPixelCount(connected_pixels, True)
edges = connected.gte(edge_length)
edgeBuffer = edges.focal_max(smooth_edges, 'square', 'meters')
# mask out areas to get histogram for Otsu
histogram_image = img.updateMask(edgeBuffer)
histogram = histogram_image.reduceRegion(ee.Reducer.histogram(255, 2)\
.combine('mean', None, True)\
.combine('variance', None,True),img.geometry(),reductionScale,bestEffort=True,
tileScale=16)
threshold = otsu(histogram.get(histBand.cat('_histogram')))
water = ee.Image(ee.Algorithms.If(invert,img.gt(threshold),img.lt(threshold)))
return water.rename('water').uint8()
if band is None:
collection = collection.select([0])
histBand = ee.String(ee.Image(collection.first()).bandNames().get(0))
else:
histBand = ee.String(band)
collection = collection.select(histBand)
return collection.map(applyEdge)
def bootstrapOtsu(collection,target_date, reductionPolygons,
neg_buffer=-1500, # negative buffer for masking potential bad data
upper_threshold=-14, # upper limit for water threshold
canny_threshold=7, # threshold for canny edge detection
canny_sigma=1, # sigma value for gaussian filter
canny_lt=7, # lower threshold for canny detection
smoothing=100, # amount of smoothing in meters
connected_pixels=200, # maximum size of the neighborhood in pixels
edge_length=50, # minimum length of edges from canny detection
smooth_edges=100,
qualityBand=None,
reverse=False,
reductionScale=90):
tDate = ee.Date(target_date)
targetColl = collection.filterDate(tDate,tDate.advance(1,'day'))
nImgs = targetColl.size().getInfo()
if nImgs <= 0:
raise EEException('Selected date has no imagery, please try processing another date')
collGeom = targetColl.geometry()
polygons = reductionPolygons.filterBounds(collGeom)
nPolys = polygons.size().getInfo()
if nPolys > 0:
ids = ee.List(polygons.aggregate_array('id'))
random_ids = []
for i in range(3):
random_ids.append(random.randint(0, ids.size().subtract(1).getInfo()))
random_ids = ee.List(random_ids)
def getRandomIds(i):
return ids.get(i)
ids = random_ids.map(getRandomIds)
polygons = polygons.filter(ee.Filter.inList('id', ids))
if qualityBand == None:
target = targetColl.mosaic().set('system:footprint', collGeom.dissolve())
target = target.clip(target.geometry().buffer(neg_buffer))
smoothed = target.focal_median(smoothing, 'circle', 'meters')
histBand = ee.String(target.bandNames().get(0))
else:
target = targetColl.qualityMosaic(qualityBand).set('system:footprint', collGeom.dissolve())
target = target.clip(target.geometry().buffer(neg_buffer))
smoothed = target.focal_median(smoothing, 'circle', 'meters')
histBand = ee.String(qualityBand)
canny = ee.Algorithms.CannyEdgeDetector(smoothed,canny_threshold,canny_sigma)
connected = canny.mask(canny).lt(canny_lt).connectedPixelCount(connected_pixels, True)
edges = connected.gte(edge_length)
edgeBuffer = edges.focal_max(smooth_edges, 'square', 'meters')
histogram_image = smoothed.mask(edgeBuffer)
histogram = histogram_image.reduceRegion(ee.Reducer.histogram(255, 2),polygons.geometry(),reductionScale,bestEffort=True)
threshold = ee.Number(otsu_function(histogram.get(histBand))).min(upper_threshold)
else:
threshold = upper_threshold
water = smoothed.lt(threshold).clip(geeutils.LAND.geometry())
return water.rename('water').uint8()\
.copyProperties(img)\
.set('system:time_start',img.get('system:time_start'))
def otsu(histogram):
counts = ee.Array(ee.Dictionary(histogram).get('histogram'))
means = ee.Array(ee.Dictionary(histogram).get('bucketMeans'))
size = means.length().get([0])
total = counts.reduce(ee.Reducer.sum(), [0]).get([0])
sums = means.multiply(counts).reduce(ee.Reducer.sum(), [0]).get([0])
mean = sums.divide(total)
indices = ee.List.sequence(1, size)
#Compute between sum of squares, where each mean partitions the data.
def bss_function(i):
aCounts = counts.slice(0, 0, i)
aCount = aCounts.reduce(ee.Reducer.sum(), [0]).get([0])
aMeans = means.slice(0, 0, i)
aMean = aMeans.multiply(aCounts).reduce(ee.Reducer.sum(), [0]).get([0]).divide(aCount)
bCount = total.subtract(aCount)
bMean = sums.subtract(aCount.multiply(aMean)).divide(bCount)
return aCount.multiply(aMean.subtract(mean).pow(2)).add(
bCount.multiply(bMean.subtract(mean).pow(2)))
bss = indices.map(bss_function)
output = means.sort(bss).get([-1])
return output