TABLE 3.

Results for scene recognition among three classifiers over UCM dataset.

Categories	ANN			XGBoost				AlexNet
	Pn	Rc	F1 Scr	Pn	Rc	F1 Scr	Pn	Rc	F1 Scr
AG	0.755	0.788	0.755	0.799	0.899	0.817	0.901	0.977	0.937
AP	0.711	0.744	0.875	0.815	0.875	0.844	0.883	0.965	0.922
BB	0.783	0.711	0.746	0.841	0.819	0.747	0.995	0.951	0.972
BH	0.792	0.658	0.737	0.844	0.889	0.844	0.986	0.937	0.960
BD	0.701	0.725	0.758	0.889	0.839	0.889	0.967	0.903	0.933
CH	0.745	0.715	0.875	0.872	0.921	0.872	0.977	0.839	0.903
DN	0.799	0.791	0.795	0.886	0.938	0.886	0.872	0.921	0.895
FR	0.783	0.711	0.746	0.985	0.954	0.985	0.886	0.938	0.911
FW	0.771	0.792	0.781	0.901	0.859	0.901	0.985	0.954	0.969
GC	0.730	0.717	0.961	0.883	0.965	0.883	0.925	0.917	0.969
HR	0.755	0.788	0.755	0.879	0.851	0.879	0.936	0.977	0.937
IN	0.711	0.744	0.875	0.986	0.937	0.986	0.871	0.871	0.922
MR	0.783	0.711	0.746	0.967	0.809	0.967	0.995	0.951	0.972
MH	0.792	0.658	0.737	0.845	0.856	0.850	0.956	0.879	0.96
OP	0.701	0.725	0.758	0.879	0.851	0.864	0.891	0.903	0.933
PN	0.874	0.845	0.859	0.986	0.937	0.960	0.819	0.916	0.859
RV	0.869	0.829	0.903	0.967	0.809	0.880	0.977	0.921	0.903
RW	0.872	0.851	0.895	0.844	0.855	0.850	0.887	0.911	0.895
SP	0.886	0.918	0.911	0.899	0.839	0.867	0.793	0.935	0.911
SN	0.965	0.934	0.969	0.872	0.875	0.873	0.799	0.916	0.895
TC	0.901	0.859	0.937	0.886	0.913	0.899	0.855	0.891	0.911
Mean	0.893	0.881	0.922	0.922	0.911	0.915	0.923	0.915	0.946

^a AG, agriculture; AP, airplane; BB, baseball diamond; BH, beach; BD, building; CH, chaparral; DN, dense residential; FR, forest; FW, Freeway GC, golf course; HR, harbor; IN, intersection; MR, medium residential; MH, mobile home park; OP, overpass; PN, parking; RV, river; RW, runway; Sp, Sparse Residential; SN, storage tank; TC, tennis court; ANN, Artificial Neural Network; XGBoost, eXtreme Gradient Bossting. Bold values indicates proposed results (highlighed).