Table 3.

Parameter configuration.

Parameter configuration	Values
Input image size	100 ∗ 100
Pooling layer	Max-polling
Dropout layer	Dropout
Fully connected layer	Flatten
Fully connected layer	Dense
Filter numbers	16, 32, 64
Filter size	3 ∗ 3
Optimizer	Adam
Learning rate	0.001
Loss function	Categorical crossentropy