-
Notifications
You must be signed in to change notification settings - Fork 11
/
Copy pathanomaly_detection.json
316 lines (316 loc) · 15.6 KB
/
anomaly_detection.json
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
310
311
312
313
314
315
316
{
"id": "anomaly_detection",
"summary": "Regional Benchmarking using CropSAR",
"description": "With the openEO-based Regional Benchmarking service in anomaly identification, you can check the crop growth on a field and compare it with a similar fields. It gives you an idea of whether your field is performing better or worse than other fields.",
"categories": [
"cubes"
],
"parameters": [
{
"name": "date",
"description": "Left-closed temporal interval, i.e. an array with exactly two elements:\n\n1. The first element is the start of the temporal interval. The specified instance in time is **included** in the interval.\n2. The second element is the end of the temporal interval. The specified instance in time is **excluded** from the interval.\n\nThe specified temporal strings follow [RFC 3339](https://www.rfc-editor.org/rfc/rfc3339.html). Also supports open intervals by setting one of the boundaries to `null`, but never both.",
"schema": {
"type": "array",
"subtype": "temporal-interval",
"minItems": 2,
"maxItems": 2,
"items": {
"anyOf": [
{
"type": "string",
"format": "date-time",
"subtype": "date-time"
},
{
"type": "string",
"format": "date",
"subtype": "date"
},
{
"type": "string",
"subtype": "year",
"minLength": 4,
"maxLength": 4,
"pattern": "^\\d{4}$"
},
{
"type": "null"
}
]
},
"examples": [
[
"2015-01-01T00:00:00Z",
"2016-01-01T00:00:00Z"
],
[
"2015-01-01",
"2016-01-01"
]
]
}
},
{
"name": "polygon",
"description": "Loaded Feature collection object of the field geometries",
"schema": {
"type": "object",
"subtype": "geojson"
}
},
{
"name": "biopar_type",
"description": "BIOPAR type [FAPAR,FCOVER] used to calculate the CropSAR curve on. As default the FAPAR is used",
"schema": {
"type": "string"
},
"default": "FAPAR",
"optional": true
}
],
"returns": {
"description": "A datacube with the CropSAR fAPAR values of the individual fields and the regional average fAPAR value for the specified time period",
"schema": {
"type": "object"
}
},
"process_graph": {
"loadcollection1": {
"process_id": "load_collection",
"arguments": {
"bands": [
"VH",
"VV"
],
"id": "SENTINEL1_GAMMA0_SENTINELHUB",
"properties": {
"polarization": {
"process_graph": {
"eq1": {
"process_id": "eq",
"arguments": {
"x": {
"from_parameter": "value"
},
"y": "DV"
},
"result": true
}
}
}
},
"spatial_extent": null,
"temporal_extent": null
}
},
"sarbackscatter1": {
"process_id": "sar_backscatter",
"arguments": {
"coefficient": "gamma0-terrain",
"contributing_area": false,
"data": {
"from_node": "loadcollection1"
},
"elevation_model": null,
"ellipsoid_incidence_angle": false,
"local_incidence_angle": false,
"mask": false,
"noise_removal": true
}
},
"loadcollection2": {
"process_id": "load_collection",
"arguments": {
"bands": [
"B03",
"B04",
"B08",
"sunAzimuthAngles",
"sunZenithAngles",
"viewAzimuthMean",
"viewZenithMean",
"SCL"
],
"id": "SENTINEL2_L2A_SENTINELHUB",
"spatial_extent": null,
"temporal_extent": null
}
},
"maskscldilation1": {
"process_id": "mask_scl_dilation",
"arguments": {
"data": {
"from_node": "loadcollection2"
},
"scl_band_name": "SCL"
}
},
"resamplecubespatial1": {
"process_id": "resample_cube_spatial",
"arguments": {
"data": {
"from_node": "maskscldilation1"
},
"method": "near",
"target": {
"from_node": "sarbackscatter1"
}
}
},
"reducedimension1": {
"process_id": "reduce_dimension",
"arguments": {
"data": {
"from_node": "resamplecubespatial1"
},
"dimension": "bands",
"reducer": {
"process_graph": {
"runudf1": {
"process_id": "run_udf",
"arguments": {
"context": {
"biopar": {
"from_parameter": "biopar_type"
}
},
"data": {
"from_parameter": "data"
},
"runtime": "Python",
"udf": "import numpy as np\nfrom typing import Dict\nfrom openeo.udf.xarraydatacube import XarrayDataCube\nimport tensorflow as tf\nfrom biopar.bioparnnw import BioParNNW\n\n\nbiopar_version = '3band'\n\ndef apply_datacube(cube: XarrayDataCube, context: Dict) -> XarrayDataCube:\n valid_biopars= ['FAPAR','LAI','FCOVER','CWC','CCC']\n biopar = context.get(\"biopar\", \"FAPAR\")\n if biopar not in valid_biopars:\n biopar = 'FAPAR'\n\n ds = cube.get_array()\n ds_date = ds\n\n from numpy import cos, radians\n ### LOAD THE DIFFERENT REQUIRED BANDS FOR THE 8-BAND FAPAR\n scaling_bands = 0.0001\n\n saa = ds_date.sel(bands='sunAzimuthAngles')\n sza = ds_date.sel(bands=\"sunZenithAngles\")\n vaa = ds_date.sel(bands=\"viewAzimuthMean\")\n vza = ds_date.sel(bands=\"viewZenithMean\")\n\n B03 = ds_date.sel(bands='B03') * scaling_bands\n B04 = ds_date.sel(bands='B04') * scaling_bands\n B8 = ds_date.sel(bands='B08') * scaling_bands\n\n g1 = cos(radians(vza))\n g2 = cos(radians(sza))\n g3 = cos(radians(saa - vaa))\n\n #### FLATTEN THE ARRAY ####\n flat = list(map(lambda arr: arr.flatten(),\n [B03.values, B04.values,B8.values, g1.values, g2.values, g3.values]))\n bands = np.array(flat)\n\n #### CALCULATE THE BIOPAR BASED ON THE BANDS #####\n image = BioParNNW(version='3band', parameter=biopar, singleConfig = True).run(bands, output_scale=1,\n output_dtype=tf.dtypes.float32,\n minmax_flagging=False) # netcdf algorithm\n as_image = image.reshape((g1.shape))\n ## set nodata to nan\n as_image[np.where(np.isnan(B03))] = np.nan\n xr_biopar = vza.copy()\n xr_biopar.values = as_image\n\n return XarrayDataCube(xr_biopar) # xarray.DataArray(as_image,vza.dims,vza.coords)\n\n\n"
},
"result": true
}
}
}
}
},
"adddimension1": {
"process_id": "add_dimension",
"arguments": {
"data": {
"from_node": "reducedimension1"
},
"label": "band_0",
"name": "bands",
"type": "bands"
}
},
"mergecubes1": {
"process_id": "merge_cubes",
"arguments": {
"cube1": {
"from_node": "sarbackscatter1"
},
"cube2": {
"from_node": "adddimension1"
}
}
},
"arrayelement1": {
"process_id": "array_element",
"arguments": {
"data": {
"from_parameter": "date"
},
"index": 0
}
},
"dateshift1": {
"process_id": "date_shift",
"arguments": {
"date": {
"from_node": "arrayelement1"
},
"unit": "day",
"value": -90
}
},
"arrayelement2": {
"process_id": "array_element",
"arguments": {
"data": {
"from_parameter": "date"
},
"index": 1
}
},
"dateshift2": {
"process_id": "date_shift",
"arguments": {
"date": {
"from_node": "arrayelement2"
},
"unit": "day",
"value": 90
}
},
"filtertemporal1": {
"process_id": "filter_temporal",
"arguments": {
"data": {
"from_node": "mergecubes1"
},
"extent": [
{
"from_node": "dateshift1"
},
{
"from_node": "dateshift2"
}
]
}
},
"vectorbuffer1": {
"process_id": "vector_buffer",
"arguments": {
"distance": -10,
"geometry": {
"from_parameter": "polygon"
},
"unit": "meter"
}
},
"aggregatespatial1": {
"process_id": "aggregate_spatial",
"arguments": {
"data": {
"from_node": "filtertemporal1"
},
"geometries": {
"from_node": "vectorbuffer1"
},
"reducer": {
"process_graph": {
"mean1": {
"process_id": "mean",
"arguments": {
"data": {
"from_parameter": "data"
}
},
"result": true
}
}
}
}
},
"runudf2": {
"process_id": "run_udf",
"arguments": {
"context": {
"date": {
"from_parameter": "date"
}
},
"data": {
"from_node": "aggregatespatial1"
},
"runtime": "Python",
"udf": "import logging\nfrom openeo.udf.udf_data import UdfData\nfrom openeo.udf.structured_data import StructuredData\nfrom openeo.rest.conversions import timeseries_json_to_pandas\nimport pandas as pd\n#import sys\n#sys.path.append(r'/data/users/Public/bontek/Nextland/Anomaly_detection/cropsar-1.4.7-py3-none-any.whl') #TODO TO remove\nfrom cropsar.preprocessing.retrieve_timeseries_openeo import run_cropsar_dataframes\n\nlogger = logging.getLogger(\"nextland_services.cropsar\")\n\n#calculate the cropsar curve for each field and the regional average of all the input fields\n######## FUNCTIONS ################\ndef get_cropsar_TS(ts_df, unique_ids_fields, metrics_order, time_range, Spark=True):\n index_fAPAR = metrics_order.index('FAPAR')\n column_indices = ts_df.columns.get_level_values(1)\n indices = column_indices.isin([index_fAPAR])\n\n df_S2 = ts_df.loc[:, indices].sort_index().T\n if(df_S2.empty):\n raise ValueError(\"Received an empty Sentinel-2 input dataframe while trying to compute cropSAR!\")\n\n df_VHVV = ts_df.loc[:, column_indices.isin([0, 1])].sort_index().T\n\n cropsar_df, cropsar_df_q10, cropsar_df_q90 = run_cropsar_dataframes(\n df_S2, df_VHVV, None, scale=1, offset=0, date_range=time_range,\n )\n\n if (len(cropsar_df.index) == 0):\n logger.warning(\"CropSAR returned an empty dataframe. For input, Sentinel-2 input: \")\n logger.warning(str(df_S2.to_json(indent=2)))\n logger.warning(\"Sentinel-1 VH-VV: \")\n logger.warning(str(df_VHVV.to_json(indent=2)))\n\n cropsar_df = cropsar_df.rename(\n columns=dict(zip(list(cropsar_df.columns.values), [str(item) + '_cropSAR' for item in unique_ids_fields])))\n cropsar_df = cropsar_df.round(decimals=3)\n cropsar_df.index = pd.to_datetime(cropsar_df.index).date\n cropsar_df = cropsar_df.loc[pd.to_datetime(time_range[0], format = '%Y-%m-%d').date() :pd.to_datetime(time_range[1], format = '%Y-%m-%d').date()]\n return cropsar_df\n\n\ndef udf_anomaly_detection(udf_data:UdfData):\n ## constants\n user_context = udf_data.user_context\n time_range = user_context.get('date')\n columns_order = ['VH', 'VV', 'FAPAR']\n\n ## load the TS\n ts_dict = udf_data.get_structured_data_list()[0].data\n if not ts_dict: # workaround of ts_dict is empty\n return\n TS_df = timeseries_json_to_pandas(ts_dict)\n TS_df.index = pd.to_datetime(TS_df.index).date\n\n if not isinstance(TS_df.columns, pd.MultiIndex):\n TS_df.columns = pd.MultiIndex.from_product([[0], TS_df.columns])\n\n amount_fields = next(iter(ts_dict.values()))\n unique_ids_fields = ['Field_{}'.format(str(p)) for p in range(len(amount_fields))]\n\n logger.info(\"CropSAR input dataframe:\\n\" + str(TS_df.describe()))\n logger.info(\"Input time range: \" + str(time_range))\n # logger.warning(str(TS_df.to_json(indent=2,date_format='iso',double_precision=5)))\n\n ts_df_cropsar = get_cropsar_TS(TS_df, unique_ids_fields, columns_order, time_range)\n\n # calculate the regional greenness curve\n ts_df_cropsar['Regional_average'] = ts_df_cropsar.mean(axis=1)\n ts_df_cropsar = ts_df_cropsar.round(decimals=3)\n ts_df_cropsar.index = ts_df_cropsar.index.astype(str)\n\n udf_data.set_structured_data_list([StructuredData(description= 'anomaly_detection', data = ts_df_cropsar.to_dict(), type = \"dict\")])\n\n\n return udf_data"
},
"result": true
}
}
}