forked from NRCan/geo-deep-learning
-
Notifications
You must be signed in to change notification settings - Fork 0
/
evaluate_segmentation.py
161 lines (137 loc) · 7.48 KB
/
evaluate_segmentation.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
import itertools
import time
from collections import defaultdict
from pathlib import Path
from typing import Sequence
import numpy as np
import pandas as pd
import rasterio
from mlflow import log_metrics
from shapely.geometry import Polygon
from tqdm import tqdm
import geopandas as gpd
from dataset.aoi import aois_from_csv
from utils.geoutils import vector_to_raster
from utils.metrics import ComputePixelMetrics
from utils.utils import get_key_def
from utils.logger import get_logger
logging = get_logger(__name__)
def metrics_per_tile(label_arr: np.ndarray, pred_img: np.ndarray, input_image: rasterio.DatasetReader,
chunk_size: int, gpkg_name: str, num_classes: int) -> gpd.GeoDataFrame:
"""
Compute metrics for each tile processed during inference
@param label_arr: numpy array of label
@param pred_img: numpy array of prediction
@param input_image: Rasterio file handle holding the (already opened) input raster
@param chunk_size: tile size for per-tile metrics
@param gpkg_name: name of geopackage
@param num_classes: number of classes
@return:
"""
xmin, ymin, xmax, ymax = input_image.bounds # left, bottom, right, top
xres, yres = (abs(input_image.transform.a), abs(input_image.transform.e))
mx = chunk_size * xres
my = chunk_size * yres
h, w = input_image.shape
feature = defaultdict(list)
cnt = 0
for row, col in tqdm(itertools.product(range(0, h, chunk_size), range(0, w, chunk_size)), leave=False,
desc="Calculating metrics per tile"):
label = label_arr[row:row + chunk_size, col:col + chunk_size]
pred = pred_img[row:row + chunk_size, col:col + chunk_size]
pixelMetrics = ComputePixelMetrics(label.flatten(), pred.flatten(), num_classes)
eval = pixelMetrics.update(pixelMetrics.iou)
feature['id_image'].append(gpkg_name)
for c_num in range(num_classes):
feature['L_count_' + str(c_num)].append(int(np.count_nonzero(label == c_num)))
feature['P_count_' + str(c_num)].append(int(np.count_nonzero(pred == c_num)))
feature['IoU_' + str(c_num)].append(eval['iou_' + str(c_num)])
feature['mIoU'].append(eval['macro_avg_iou'])
logging.debug(eval['macro_avg_iou'])
x_1, y_1 = (xmin + (col * xres)), (ymax - (row * yres))
x_2, y_2 = (xmin + ((col * xres) + mx)), y_1
x_3, y_3 = x_2, (ymax - ((row * yres) + my))
x_4, y_4 = x_1, y_3
geom = Polygon([(x_1, y_1), (x_2, y_2), (x_3, y_3), (x_4, y_4)])
feature['geometry'].append(geom)
feature['length'].append(geom.length)
feature['pointx'].append(geom.centroid.x)
feature['pointy'].append(geom.centroid.y)
feature['area'].append(geom.area)
cnt += 1
gdf = gpd.GeoDataFrame(feature, crs=input_image.crs.to_epsg())
return gdf
def main(params):
"""
Computes benchmark metrics from inference and ground truth and write results to a gpkg.
@param params:
@return:
"""
start_seg = time.time()
state_dict = get_key_def('state_dict_path', params['inference'], to_path=True, validate_path_exists=True)
bands_requested = get_key_def('bands', params['dataset'], default=None, expected_type=Sequence)
num_bands = len(bands_requested)
working_folder = state_dict.parent.joinpath(f'inference_{num_bands}bands')
raw_data_csv = get_key_def('raw_data_csv', params['inference'], default=working_folder,
expected_type=str, to_path=True, validate_path_exists=True)
num_classes = len(get_key_def('classes_dict', params['dataset']).keys())
single_class_mode = True if num_classes == 1 else False
threshold = 0.5
debug = get_key_def('debug', params, default=False, expected_type=bool)
# benchmark (ie when gkpgs are inputted along with imagery)
out_gpkg = get_key_def('out_benchmark_gpkg', params['inference'], default=working_folder/"benchmark.gpkg",
expected_type=str)
chunk_size = get_key_def('chunk_size', params['inference'], default=512, expected_type=int)
dontcare = get_key_def("ignore_index", params["dataset"], -1)
attribute_field = get_key_def('attribute_field', params['dataset'], None, expected_type=str)
attr_vals = get_key_def('attribute_values', params['dataset'], None, expected_type=Sequence)
# Assert that all values are integers (ex.: to benchmark single-class model with multi-class labels)
if attr_vals:
for item in attr_vals:
if not isinstance(item, int):
raise ValueError(f'\nValue "{item}" in attribute_values is {type(item)}, expected int.')
list_aois = aois_from_csv(csv_path=raw_data_csv, bands_requested=bands_requested)
# VALIDATION: anticipate problems with imagery and label (if provided) before entering main for loop
for aoi in tqdm(list_aois, desc='Validating ground truth'):
if aoi.label is None:
raise ValueError(f"No ground truth was inputted to evaluate with")
logging.info('\nSuccessfully validated label data for benchmarking')
gdf_ = []
gpkg_name_ = []
for aoi in tqdm(list_aois, desc='Evaluating from input list', position=0, leave=True):
Path.mkdir(working_folder / aoi.raster_name.parent.name, parents=True, exist_ok=True)
inference_image = working_folder / aoi.raster_name.parent.name / f"{aoi.raster_name.stem}_inference.tif"
if not inference_image.is_file():
raise FileNotFoundError(f"Couldn't locate inference to evaluate metrics with. Make inferece has been run "
f"before you run evaluate mode.")
pred = rasterio.open(inference_image).read()[0, ...]
logging.info(f'\nBurning label as raster: {aoi.label}')
raster = rasterio.open(aoi.raster_name, 'r')
logging.info(f'\nReading image: {raster.name}')
inf_meta = raster.meta
label = vector_to_raster(vector_file=aoi.label,
input_image=raster,
out_shape=(inf_meta['height'], inf_meta['width']),
attribute_name=attribute_field,
fill=0, # background value in rasterized vector.
attribute_values=attr_vals)
if debug:
logging.debug(f'\nUnique values in loaded label as raster: {np.unique(label)}\n'
f'Shape of label as raster: {label.shape}')
gdf = metrics_per_tile(label_arr=label, pred_img=pred, input_image=raster, chunk_size=chunk_size,
gpkg_name=aoi.label.stem, num_classes=num_classes)
gdf_.append(gdf.to_crs(4326))
gpkg_name_.append(aoi.label.stem)
if 'tracker_uri' in locals():
pixelMetrics = ComputePixelMetrics(label, pred, num_classes)
log_metrics(pixelMetrics.update(pixelMetrics.iou))
log_metrics(pixelMetrics.update(pixelMetrics.dice))
if not len(gdf_) == len(gpkg_name_):
raise logging.critical(ValueError('\nbenchmarking unable to complete'))
all_gdf = pd.concat(gdf_) # Concatenate all geo data frame into one geo data frame
all_gdf.reset_index(drop=True, inplace=True)
gdf_x = gpd.GeoDataFrame(all_gdf, crs=4326)
gdf_x.to_file(out_gpkg, driver="GPKG", index=False)
logging.info(f'\nSuccessfully wrote benchmark geopackage to: {out_gpkg}')
end_seg_ = time.time() - start_seg
logging.info('Benchmark operation completed in {:.0f}m {:.0f}s'.format(end_seg_ // 60, end_seg_ % 60))