The Potential Application of Innovative Methods in Neural Networks for Surface Crack Recognition of Unshelled Hazelnut

Table 1

The parameters of DCNN.

	Inception-V3	VGG-19	ResNet-50	Propose model
Input size
Kernel size	, , , , ,, and		, , and
Pooling size	, , and		and
Number of convolution layers	159	16	50	3
Activation function	ReLU	ReLU	ReLU	ReLU