Background Modeling with Extracted Dynamic Pixels for Pumping Unit Surveillance

Table 3

Comparisons of BOWED with nine other methods on the pumping unit monitoring videos.

Method	video	Precision	Recall	F-Measure	FPR	FNR	PWC
BOWED	1		0.6507	0.7842		0.3493
	2		0.6354	0.7752		0.3646	0.0420
	3		0.8366		0.0025	0.1634
ViBe [10]	1	0.5146	0.6441	0.5721	0.0213	0.3559	0.0327
	2	0.9319	0.6517	0.7670	0.0061	0.3483	0.0451
	3	0.5363	0.7331	0.6195	0.0400	0.2669	0.0535
GMM [35]	1	0.7823	0.6027	0.6808	0.0059	0.3973	0.0192
	2	0.6197	0.1423	0.2315	0.0112	0.8577	0.1077
	3	0.7413	0.3894	0.5106	0.0086	0.6106	0.0443
KDE [11]	1	0.0749	0.9401	0.1387	0.4079	0.0599	0.3961
	2	0.3682	0.8760	0.5184	0.1933	0.1240	0.1854
	3	0.3609	0.9866	0.5285	0.1103	0.0134	0.1045
PBAS [13]	1	0.7390	0.8783		0.0109	0.1217	0.0146
	2	0.7939	0.7402	0.7661	0.0247	0.2598	0.0515
	3	0.6261	0.7502	0.6826	0.02883	0.2498	0.0414
SuBSENSE [27]	1	0.5196	0.9370	0.6685	0.0304	0.0630	0.0315
	2	0.8010	0.8019	0.8015	0.0256	0.1981	0.0453
	3	0.5585	0.9782	0.7111	0.0488	0.0218	0.0472
GMG [36]	1	0.3029		0.4632	0.0796		0.0774
	2	0.4337		0.5959	0.1598		0.1471
	3	0.5679		0.7220	0.0476		0.0453
LOBSTER [28]	1	0.4180	0.8802	0.5669	0.0430	0.1198	0.0456
	2	0.9719	0.7194		0.0027	0.2806
	3	0.5613	0.9472	0.7049	0.0467	0.0528	0.0471
T2-FGMM with MRF [37]	1	0.4638	0.0123	0.0239	0.0005	0.9877	0.0340
	2	0.4340	0.0020	0.0040	0.0005	0.9980	0.1140
	3	0.7001	0.0260	0.0501		0.9740	0.0585
IMBS [38]	1	0.4509	0.8154	0.5806	0.0349	0.1846	0.0399
	2	0.8682	0.7479	0.8036	0.0146	0.2521	0.0417
	3	0.5294	0.9302	0.6747	0.0522	0.0698	0.0533