Table 3.

Quantitative comparison of different detection methods.

Dataset	Metric	FCN	UNet	Mobile-Unet	LSTS	RPDNet
FI	IoU	0.645	0.687	0.727	0.650	0.766
	Recall	0.771	0.800	0.792	0.811	0.833
	Precision	0.798	0.822	0.898	0.766	0.905
	F1-Measure	0.784	0.809	0.842	0.788	0.867
TILDA	IoU	0.719	0.687	0.752		0.831
	Recall	0.831	0.784	0.833		0.886
	Precision	0.842	0.849	0.886		0.930
	F1-Measure	0.836	0.815	0.859		0.907