Table 1 Training and test sample numbers in the WHU-HiLongKou dataset, theWHU-HiHanChuan dataset, and the WHU-HiHongHu dataset.

From: CMTNet: a hybrid CNN-transformer network for UAV-based hyperspectral crop classification in precision agriculture

NO.

WHU-Hi LongKou

WHU-Hi HanChuan

WHU-Hi HongHu

Class

Training

Test

Class

Training

Test

Class

Training

Test

1

Corn

172

34339

Strawberry

223

44521

Red roof

70

13971

2

Cotton

41

8333

Cowpea

113

22640

Road

17

3495

3

Sesame

15

3016

Soybean

51

10236

Bare soil

109

21712

4

Broad-leaf soybean

316

62896

Sorghum

26

5327

Cotton

816

162469

5

Narrow-leaf soybean

20

4131

Water spinach

6

1194

Cotton firewood

31

6187

6

Rice

59

11795

Watermelon

22

4511

Rape

222

44335

7

Water

335

66721

Greens

29

5974

Chinese cabbage

120

23983

8

Roads and houses

35

7089

Trees

89

17889

Pakchoi

20

4034

9

Mixed weed

26

5203

Grass

47

9422

Cabbage

54

10765

10

   

Red roof

52

10464

Tuber mustard

61

12333

11

   

Gray roof

84

16827

Brassica parachinensis

55

10960

12

   

Plastic

18

3661

Brassica chinensis

44

8910

13

   

Bare soil

45

9071

Small Brassica chinensis

112

22395

14

   

Road

92

18468

Lactuca sativa

36

7320

14

   

Bright object

5

1131

Celtuce

5

997

16

   

Water

377

75024

Film covered lettuce

36

7226

17

      

Romaine lettuce

15

2995

18

      

Carrot

16

3201

19

      

White radish

43

8669

20

      

Garlic sprout

17

3469

21

      

Broad bean

6

1322

22

      

Tree

20

4020

/

Total

1019

203523

Total

1289

256241

Total

1925

384678