Table 11.

Performance results comparison among the suggested model and the other traditional pre-trained models.

Model	Class	Precision (%)	Recall (%)	F1-Score (%)	Accuracy (%)
VGG19	4	90	88.5	89	88.98
ResNet50	4	88	87.25	87.5	87.5
InceptionV3	4	87.5	87.25	87.5	87.39
VGG19	3	93.25	94	93	94.35
ResNet50	3	95	93.5	94	94.88
InceptionV3	3	93	92.25	92.75	93.27
VGG19	2	99	98.5	98	98.5
ResNet50	2	99	98	98	98.33
InceptionV3	2	0.98	97	98	97.67
Proposed model (on Dataset-1)	4	90.05	89.85	89.65	89.76
	3	0.97	89.85	0.9633	97.22
	2	0.99	89.85	0.99	99.13
Proposed model (on Dataset-2)	4	99.5	99.4	99.5	99.46
Proposed model (on Dataset-3)	2	98.4	98.4	98.3	98.26