Skip to main content

Table 1 Size and number of training images used in each experiment

From: Identifying the vegetation type in Google Earth images using a convolutional neural network: a case study for Japanese bamboo forests

Zoom level

1/500

1/1000

1/2500

Chopping size (pix)

28

56

84

28

14

Sanyo-Onoda

67398

16207

6866

16000

16000

Ide

79043

18653

2836

18000

18000

Isumi

64106

15083

2837

15000

15000