Table 3.

Segmentation results of three different models.

Model	Type	IoU	Accuracy	F1-score	Precision	Recall	Data loading time	Model run time
DeepLabv3+	Kenaf	78.35	84.34	87.86	91.7	84.34	0.258	1.707
	Background	86.69	95.21	92.87	90.65	95.21
	Average	82.52	89.77	90.37	91.17	89.77
FCN	Kenaf	76.71	88.13	86.82	85.54	88.13	0.204	0.566
	Background	84.38	90.66	91.53	92.41	90.66
	Average	80.54	89.39	89.17	88.98	89.39
U-Net	Kenaf	81.3	92.33	89.68	87.19	92.33	0.244	0.653
	Background	87.29	91.49	93.21	95.0	91.49
	Average	84.29	91.91	91.45	91.1	91.91