Skip to main content

Table 2 Architecture of the network used in this study

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

Layer name

Filter size

Number of output neurons

Convolution 1

5 × 5

20

Pooling 1

2 × 2

 

Convolution 2

5 × 5

50

Pooling 2

2 × 2

 

Fully connected 1

500

Fully connected 2

 

2

Back to article page

BMC Ecology

ISSN: 1472-6785

Contact us