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

Table 5

Classification performances of DCNN models used in this study.

Models	Precision (%)	Recall (%)	Accuracy (%)	˗score (%)
Test data
ResNet-50	73.24	71.36	71.92	72.28
VGG-19	96.32	96.40	96.32	96.35
Inception-V3	97.71	98.12	98.15	97.91
Proposed model	96.62	97.07	97.85	96.84
Validation data
ResNet-50	72.23	73.81	72.12	73.01
VGG˗19	98.21	97.85	96.78	98.03
Inception-V3	97.94	98.02	98.41	97.98
Proposed model	97.38	97.77	98.88	97.57