Multiclass Dataset statistics#

[1]:

from aitlas.datasets import (
    AIDDataset,
    EurosatDataset,
    PatternNetDataset,
    Resisc45Dataset,
    RSD46WHUDataset,
    RSICB256Dataset,
    RSSCN7Dataset,
    SAT6Dataset,
    SiriWhuDataset,
    UcMercedDataset,
    WHURS19Dataset,
    CLRSDataset,
    BrazilianCoffeeScenesDataset,
    Optimal31Dataset,
    So2SatDataset
)

AID dataset#

[2]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/AID",
    "csv_file": "/media/hdd/multi-class/AID/train.csv"
}
train_dataset = AIDDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 8000

[2]:
Label Count
0 Airport 280
1 BareLand 249
2 BaseballField 168
3 Beach 323
4 Bridge 294
5 Center 204
6 Church 190
7 Commercial 281
8 DenseResidential 328
9 Desert 241
10 Farmland 293
11 Forest 205
12 Industrial 338
13 Meadow 213
14 MediumResidential 238
15 Mountain 270
16 Park 277
17 Parking 309
18 Playground 291
19 Pond 349
20 Port 321
21 RailwayStation 201
22 Resort 227
23 River 325
24 School 243
25 SparseResidential 224
26 Square 262
27 Stadium 236
28 StorageTanks 299
29 Viaduct 321
../_images/examples_multclass_datasets_statistics_3_2.png
../_images/examples_multclass_datasets_statistics_3_3.png

Eurosat dataset#

[4]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/EuroSAT",
    "csv_file": "/media/hdd/multi-class/EuroSAT/train.csv"
}
train_dataset = EurosatDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 21600

[4]:
Label Count
0 AnnualCrop 2421
1 Forest 2382
2 HerbaceousVegetation 2394
3 Highway 2000
4 Industrial 2004
5 Pasture 1600
6 PermanentCrop 1993
7 Residential 2426
8 River 2003
9 SeaLake 2377
../_images/examples_multclass_datasets_statistics_5_2.png
../_images/examples_multclass_datasets_statistics_5_3.png

PatternNet dataset#

[5]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/PatternNet",
    "csv_file": "/media/hdd/multi-class/PatternNet/train.csv"
}
train_dataset = PatternNetDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 24320

[5]:
Label Count
0 airplane 652
1 baseball_field 642
2 basketball_court 641
3 beach 654
4 bridge 639
5 cemetery 646
6 chaparral 630
7 christmas_tree_farm 634
8 closed_road 643
9 coastal_mansion 624
10 crosswalk 636
11 dense_residential 630
12 ferry_terminal 646
13 football_field 643
14 forest 658
15 freeway 634
16 golf_course 649
17 harbor 615
18 intersection 631
19 mobile_home_park 635
20 nursing_home 640
21 oil_gas_field 635
22 oil_well 653
23 overpass 645
24 parking_lot 637
25 parking_space 656
26 railway 647
27 river 657
28 runway 651
29 runway_marking 618
30 shipping_yard 625
31 solar_panel 634
32 sparse_residential 635
33 storage_tank 634
34 swimming_pool 632
35 tennis_court 646
36 transformer_station 652
37 wastewater_treatment_plant 641
../_images/examples_multclass_datasets_statistics_7_2.png
../_images/examples_multclass_datasets_statistics_7_3.png

Resisc45 dataset#

[6]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/RESISC45",
    "csv_file": "/media/hdd/multi-class/RESISC45/train.csv"
}
train_dataset = Resisc45Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 25200

[6]:
Label Count
0 airplane 545
1 airport 555
2 baseball_diamond 561
3 basketball_court 560
4 beach 581
5 bridge 557
6 chaparral 547
7 church 559
8 circular_farmland 562
9 cloud 563
10 commercial_area 557
11 dense_residential 565
12 desert 566
13 forest 576
14 freeway 562
15 golf_course 568
16 ground_track_field 563
17 harbor 556
18 industrial_area 568
19 intersection 569
20 island 555
21 lake 565
22 meadow 551
23 medium_residential 566
24 mobile_home_park 557
25 mountain 537
26 overpass 561
27 palace 549
28 parking_lot 579
29 railway 535
30 railway_station 567
31 rectangular_farmland 566
32 river 565
33 roundabout 558
34 runway 560
35 sea_ice 549
36 ship 564
37 snowberg 558
38 sparse_residential 566
39 stadium 568
40 storage_tank 540
41 tennis_court 572
42 terrace 570
43 thermal_power_station 556
44 wetland 546
../_images/examples_multclass_datasets_statistics_9_2.png
../_images/examples_multclass_datasets_statistics_9_3.png

RSD46WHU dataset train split#

[7]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/RSD46-WHU/train",
    "csv_file": "/media/hdd/multi-class/RSD46-WHU/train/train.csv"
}
train_dataset = RSD46WHUDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 99381

[7]:
Label Count
0 Airplane 2575
1 Airport 1405
2 Artificial dense forest land 2388
3 Artificial sparse forest land 2404
4 Bare land 851
5 Basketball court 2534
6 Blue structured factory building 2611
7 Building 2940
8 Construction site 2786
9 Cross river bridge 1910
10 Crossroads 1741
11 Dense tall building 2608
12 Dock 2675
13 Fish pond 1372
14 Footbridge 2230
15 Graff 2559
16 Grassland 2408
17 Low scattered building 2039
18 Lrregular farmland 2665
19 Medium density scattered building 894
20 Medium density structured building 3015
21 Natural dense forest land 2550
22 Natural sparse forest land 2519
23 Oiltank 1368
24 Overpass 2128
25 Parking lot 2598
26 Plasticgreenhouse 863
27 Playground 2728
28 Railway 2646
29 Red structured factory building 2536
30 Refinery 2259
31 Regular farmland 2728
32 Scattered blue roof factory building 2593
33 Scattered red roof factory building 2497
34 Sewage plant-type-one 458
35 Sewage plant-type-two 364
36 Ship 2563
37 Solar power station 2579
38 Sparse residential area 2534
39 Square 2813
40 Steelsmelter 2494
41 Storage land 1798
42 Tennis court 1322
43 Thermal power plant 1074
44 Vegetable plot 2452
45 Water 2307
../_images/examples_multclass_datasets_statistics_11_2.png
../_images/examples_multclass_datasets_statistics_11_3.png

RSD46WHU dataset test split#

[10]:

test_dataset_config = {
    "data_dir": "/media/hdd/multi-class/RSD46-WHU/val",
    "csv_file": "/media/hdd/multi-class/RSD46-WHU/val/test.csv"
}
test_dataset = RSD46WHUDataset(test_dataset_config)

print(f"Total number of images: {len(test_dataset)}")
fig = test_dataset.show_batch(20, True)
fig = test_dataset.data_distribution_barchart()
test_dataset.data_distribution_table()

Total number of images: 17512

[10]:
Label Count
0 Airplane 454
1 Airport 247
2 Artificial dense forest land 420
3 Artificial sparse forest land 424
4 Bare land 150
5 Basketball court 447
6 Blue structured factory building 460
7 Building 518
8 Construction site 491
9 Cross river bridge 337
10 Crossroads 307
11 Dense tall building 460
12 Dock 472
13 Fish pond 241
14 Footbridge 393
15 Graff 451
16 Grassland 424
17 Low scattered building 359
18 Lrregular farmland 470
19 Medium density scattered building 157
20 Medium density structured building 532
21 Natural dense forest land 450
22 Natural sparse forest land 445
23 Oiltank 241
24 Overpass 375
25 Parking lot 458
26 Plasticgreenhouse 152
27 Playground 481
28 Railway 466
29 Red structured factory building 447
30 Refinery 399
31 Regular farmland 481
32 Scattered blue roof factory building 457
33 Scattered red roof factory building 440
34 Sewage plant-type-one 80
35 Sewage plant-type-two 64
36 Ship 451
37 Solar power station 454
38 Sparse residential area 447
39 Square 496
40 Steelsmelter 439
41 Storage land 316
42 Tennis court 232
43 Thermal power plant 189
44 Vegetable plot 432
45 Water 406
../_images/examples_multclass_datasets_statistics_13_2.png
../_images/examples_multclass_datasets_statistics_13_3.png

RSICB256 dataset#

[11]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/RSI-CB256",
    "csv_file": "/media/hdd/multi-class/RSI-CB256/train.csv"
}
train_dataset = RSICB256Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 19798

[11]:
Label Count
0 airplane 288
1 airport_runway 535
2 artificial_grassland 222
3 avenue 443
4 bare_land 663
5 bridge 372
6 city_building 810
7 coastline 360
8 container 521
9 crossroads 448
10 dam 260
11 desert 866
12 dry_farm 1027
13 forest 885
14 green_farmland 516
15 highway 180
16 hirst 502
17 lakeshore 363
18 mangrove 812
19 marina 297
20 mountain 670
21 parkinglot 364
22 pipeline 162
23 residents 658
24 river 440
25 river_protection_forest 441
26 sandbeach 432
27 sapling 695
28 sea 834
29 shrubwood 1058
30 snow_mountain 920
31 sparse_forest 886
32 storage_room 1042
33 stream 554
34 town 272
../_images/examples_multclass_datasets_statistics_15_2.png
../_images/examples_multclass_datasets_statistics_15_3.png

RSSCN7 dataset#

[12]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/RSSCN7",
    "csv_file": "/media/hdd/multi-class/RSSCN7/train.csv"
}
train_dataset = RSSCN7Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 2240

[12]:
Label Count
0 farm_land 305
1 forest 322
2 grass_land 323
3 industrial_region 310
4 parking_lot 348
5 residential_region 317
6 river_lake 315
../_images/examples_multclass_datasets_statistics_17_2.png
../_images/examples_multclass_datasets_statistics_17_3.png

SAT6 train split#

[13]:

train_dataset_config = {
    "mat_file": "/media/hdd/multi-class/SAT/sat-6-full.mat",
    "mode": "train"

}
train_dataset = SAT6Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 324000

[13]:
Label Count
0 barren land 73397
1 buildings 14923
2 grassland 50347
3 roads 8192
4 trees 56809
5 water bodies 120332
../_images/examples_multclass_datasets_statistics_19_2.png
../_images/examples_multclass_datasets_statistics_19_3.png

SAT6 test split#

[14]:

test_dataset_config = {
    "mat_file": "/media/hdd/multi-class/SAT/sat-6-full.mat",
    "mode": "test"

}
test_dataset = SAT6Dataset(test_dataset_config)

print(f"Total number of images: {len(test_dataset)}")
fig = test_dataset.show_batch(20, True)
fig = test_dataset.data_distribution_barchart()
test_dataset.data_distribution_table()

Total number of images: 81000

[14]:
Label Count
0 barren land 18367
1 buildings 3714
2 grassland 12596
3 roads 2070
4 trees 14185
5 water bodies 30068
../_images/examples_multclass_datasets_statistics_21_2.png
../_images/examples_multclass_datasets_statistics_21_3.png

SiriWhu dataset#

[15]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/SIRI-WHU",
    "csv_file": "/media/hdd/multi-class/SIRI-WHU/train.csv"
}
train_dataset = SiriWhuDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 1920

[15]:
Label Count
0 agriculture 157
1 commercial 163
2 harbor 158
3 idle_land 153
4 industrial 159
5 meadow 161
6 overpass 168
7 park 159
8 pond 153
9 residential 161
10 river 167
11 water 161
../_images/examples_multclass_datasets_statistics_23_2.png
../_images/examples_multclass_datasets_statistics_23_3.png

UcMerced dataset#

[16]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/UCMerced",
    "csv_file": "/media/hdd/multi-class/UCMerced/train.csv"
}
train_dataset = UcMercedDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 1680

[16]:
Label Count
0 agricultural 81
1 airplane 79
2 baseballdiamond 77
3 beach 89
4 buildings 76
5 chaparral 84
6 denseresidential 81
7 forest 81
8 freeway 79
9 golfcourse 76
10 harbor 79
11 intersection 84
12 mediumresidential 81
13 mobilehomepark 78
14 overpass 77
15 parkinglot 78
16 river 78
17 runway 82
18 sparseresidential 82
19 storagetanks 79
20 tenniscourt 79
../_images/examples_multclass_datasets_statistics_25_2.png
../_images/examples_multclass_datasets_statistics_25_3.png

WHURS19 dataset#

[17]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/WHU-RS19",
    "csv_file": "/media/hdd/multi-class/WHU-RS19/train.csv"
}
train_dataset = WHURS19Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 804

[17]:
Label Count
0 Airport 47
1 Beach 39
2 Bridge 39
3 Commercial 46
4 Desert 37
5 Farmland 40
6 Forest 41
7 Industrial 42
8 Meadow 49
9 Mountain 39
10 Park 40
11 Parking 40
12 Pond 42
13 Port 43
14 Residential 47
15 River 45
16 Viaduct 42
17 footballField 46
18 railwayStation 40
../_images/examples_multclass_datasets_statistics_27_2.png
../_images/examples_multclass_datasets_statistics_27_3.png

CLRS dataset#

[18]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/CLRS",
    "csv_file": "/media/hdd/multi-class/CLRS/train.csv"
}
train_dataset = CLRSDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 12000

[18]:
Label Count
0 airport 479
1 bare-land 477
2 beach 476
3 bridge 486
4 commercial 476
5 desert 481
6 farmland 498
7 forest 497
8 golf-course 466
9 highway 481
10 industrial 491
11 meadow 473
12 mountain 484
13 overpass 482
14 park 456
15 parking 465
16 playground 480
17 port 483
18 railway 488
19 railway-station 492
20 residential 469
21 river 476
22 runway 474
23 stadium 490
24 storage-tank 480
../_images/examples_multclass_datasets_statistics_29_2.png
../_images/examples_multclass_datasets_statistics_29_3.png

Brazilian Coffee Scenes dataset#

[19]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/BCS",
    "csv_file": "/media/hdd/multi-class/BCS/train.csv"
}
train_dataset = BrazilianCoffeeScenesDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 2876

[19]:
Label Count
0 coffee 1438
1 noncoffee 1438
../_images/examples_multclass_datasets_statistics_31_2.png
../_images/examples_multclass_datasets_statistics_31_3.png

Optimal 31 Dataset#

[20]:

train_dataset_config = {
    "data_dir": "/media/hdd/multi-class/Optimal31",
    "csv_file": "/media/hdd/multi-class/Optimal31/train.csv"
}
train_dataset = Optimal31Dataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 1860

[20]:
Label Count
0 airplane 60
1 airport 60
2 baseball_diamond 60
3 basketball_court 60
4 beach 60
5 bridge 60
6 chaparral 60
7 church 60
8 circular_farmland 60
9 commercial_area 60
10 dense_residential 60
11 desert 60
12 forest 60
13 freeway 60
14 golf_course 60
15 ground_track_field 60
16 harbor 60
17 industrial_area 60
18 intersection 60
19 island 60
20 lake 60
21 meadow 60
22 medium_residential 60
23 mobile_home_park 60
24 mountain 60
25 overpass 60
26 parking_lot 60
27 railway 60
28 rectangular_farmland 60
29 roundabout 60
30 runway 60
../_images/examples_multclass_datasets_statistics_33_2.png
../_images/examples_multclass_datasets_statistics_33_3.png

So2Sat Dataset train split#

[22]:

train_dataset_config = {
    "h5_file": "/media/hdd/multi-class/So2Sat/training.h5"
}
train_dataset = So2SatDataset(train_dataset_config)

print(f"Total number of images: {len(train_dataset)}")
fig = train_dataset.show_batch(20, True)
fig = train_dataset.data_distribution_barchart()
train_dataset.data_distribution_table()

Total number of images: 352366

[22]:
Label Count
0 Compact high_rise 5068.0
1 Compact middle_rise 24431.0
2 Compact low_rise 31693.0
3 Open high_rise 8651.0
4 Open middle_rise 16493.0
5 Open low_rise 35290.0
6 Lightweight low_rise 3269.0
7 Large low_rise 39326.0
8 Sparsely built 13584.0
9 Heavy industry 11954.0
10 Dense trees 42902.0
11 Scattered trees 9514.0
12 Bush or scrub 9165.0
13 Low plants 41377.0
14 Bare rock or paved 2392.0
15 Bare soil or sand 7898.0
16 Water 49359.0
../_images/examples_multclass_datasets_statistics_35_2.png
../_images/examples_multclass_datasets_statistics_35_3.png

So2Sat Dataset validation split#

[23]:

validation_dataset_config = {
    "h5_file": "/media/hdd/multi-class/So2Sat/validation.h5"
}
validation_dataset = So2SatDataset(validation_dataset_config)

print(f"Total number of images: {len(validation_dataset)}")
fig = validation_dataset.show_batch(20, True)
fig = validation_dataset.data_distribution_barchart()
validation_dataset.data_distribution_table()

Total number of images: 24119

[23]:
Label Count
0 Compact high_rise 256.0
1 Compact middle_rise 1254.0
2 Compact low_rise 2353.0
3 Open high_rise 849.0
4 Open middle_rise 757.0
5 Open low_rise 1906.0
6 Lightweight low_rise 474.0
7 Large low_rise 3395.0
8 Sparsely built 1914.0
9 Heavy industry 860.0
10 Dense trees 2287.0
11 Scattered trees 382.0
12 Bush or scrub 1202.0
13 Low plants 2747.0
14 Bare rock or paved 202.0
15 Bare soil or sand 672.0
16 Water 2609.0
../_images/examples_multclass_datasets_statistics_37_2.png
../_images/examples_multclass_datasets_statistics_37_3.png

So2Sat Dataset test split#

[24]:

test_dataset_config = {
    "h5_file": "/media/hdd/multi-class/So2Sat/testing.h5"
}
test_dataset = So2SatDataset(test_dataset_config)

print(f"Total number of images: {len(test_dataset)}")
fig = test_dataset.show_batch(20, True)
fig = test_dataset.data_distribution_barchart()
test_dataset.data_distribution_table()

Total number of images: 24188

[24]:
Label Count
0 Compact high_rise 266.0
1 Compact middle_rise 1262.0
2 Compact low_rise 2465.0
3 Open high_rise 857.0
4 Open middle_rise 763.0
5 Open low_rise 1936.0
6 Lightweight low_rise 503.0
7 Large low_rise 3496.0
8 Sparsely built 2013.0
9 Heavy industry 908.0
10 Dense trees 2368.0
11 Scattered trees 433.0
12 Bush or scrub 1166.0
13 Low plants 2435.0
14 Bare rock or paved 205.0
15 Bare soil or sand 572.0
16 Water 2540.0
../_images/examples_multclass_datasets_statistics_39_2.png
../_images/examples_multclass_datasets_statistics_39_3.png 
[ ]: