Method BSABU on the SBM-RGBD dataset

Contact name: Navid Dorudian

Contact email: navid.dorudian@brunel.ac.uk

Contact Institution: Brunel University

Method processing time: ~12 fps for a 640x480 video with C++ code running on a core i7-6700HQ CPU @2.6 GHz laptop

Method parameters: N = 20, noMin=3

Reference: Navid Dorudian, Stanislao Lauria, Stephen Swift, Moving Object Detection Using Adaptive Blind Update and RGB-D Camera, IEEE Sensors Journal 19(18), 2019.

Get all result files: here.

A) Sequence-by-sequence results

Name	Recall	Specificity	FPR	FNR	PWC	Precision	F-Measure
Bootstrapping
BootStrapping_ds	0.8870	0.9994	0.0006	0.1130	1.1188	0.9931	0.9371
adl24cam0	0.9251	0.9972	0.0028	0.0749	0.7125	0.9560	0.9403
bear_front	0.9312	0.9942	0.0058	0.0688	2.2406	0.9828	0.9563
fall01cam0	0.8182	0.9991	0.0009	0.1818	1.0996	0.9808	0.8921
fall20cam0	0.4736	0.9874	0.0126	0.5264	3.9100	0.6704	0.5550
ColorCamouflage
Cespatx_ds	0.8431	0.9997	0.0003	0.1569	0.3480	0.9843	0.9083
Hallway	0.8023	0.9916	0.0084	0.1977	1.8282	0.8406	0.8210
colorCam1	0.9604	0.9988	0.0012	0.0396	0.9785	0.9958	0.9778
colorCam2	0.9862	0.9789	0.0211	0.0138	1.7786	0.9749	0.9805
DepthCamouflage
DCamSeq1	0.9109	0.9982	0.0018	0.0891	0.6637	0.9683	0.9387
DCamSeq2	0.8516	0.9974	0.0026	0.1484	0.8997	0.9370	0.8923
Despatx_ds	0.9724	0.9986	0.0014	0.0276	0.3820	0.9853	0.9788
Wall	0.8551	0.9997	0.0003	0.1449	1.3230	0.9963	0.9203
IlluminationChanges
ChairBox	0.8627	0.9986	0.0014	0.1373	1.6202	0.9869	0.9206
Ls_ds	0.0000	0.9917	0.0083	0.0000	0.8342	0.0000	0.0000
TimeOfDay_ds	0.0000	0.9854	0.0146	0.0000	1.4552	0.0000	0.0000
genSeq1	0.9884	0.9863	0.0137	0.0116	1.3458	0.9121	0.9488
IntermittentMotion
Shelves	0.7055	0.9995	0.0005	0.2945	0.6218	0.9684	0.8163
Sleeping_ds	0.9563	0.9951	0.0049	0.0437	1.2674	0.9802	0.9681
abandoned1	0.9520	0.9991	0.0009	0.0480	0.2098	0.9637	0.9578
abandoned2	0.9635	0.9972	0.0028	0.0365	0.6925	0.9791	0.9712
movedBackground1	0.9138	0.9988	0.0012	0.0862	0.2498	0.9169	0.9153
movedBackground2	0.9083	0.9944	0.0056	0.0917	1.1085	0.9176	0.9129
OutOfRange
MultiPeople1	0.8498	0.9969	0.0031	0.1502	1.4618	0.9587	0.9010
MultiPeople2	0.8393	0.9963	0.0037	0.1607	1.6460	0.9533	0.8927
TopViewLab1	0.8810	0.9964	0.0036	0.1190	0.7479	0.8929	0.8869
TopViewLab2	0.9249	0.9979	0.0021	0.0751	0.4934	0.9457	0.9352
TopViewLab3	0.7712	0.9959	0.0041	0.2288	0.8860	0.8043	0.7874
Shadows
Shadows_ds	0.9265	0.9987	0.0013	0.0735	0.5732	0.9791	0.9521
fall01cam1	0.7581	0.9984	0.0016	0.2419	1.6690	0.9704	0.8512
genSeq2	0.9828	0.9963	0.0037	0.0172	0.5230	0.9706	0.9767
shadows1	0.9875	0.9984	0.0016	0.0125	0.3157	0.9900	0.9887
shadows2	0.9901	0.9979	0.0021	0.0099	0.3129	0.9871	0.9886

B) Average results for each category

Name	Recall	Specificity	FPR	FNR	PWC	Precision	F-Measure
Bootstrapping	0.8070	0.9954	0.0046	0.1930	1.8163	0.9166	0.8562
ColorCamouflage	0.8980	0.9923	0.0077	0.1020	1.2333	0.9489	0.9219
DepthCamouflage	0.8975	0.9985	0.0015	0.1025	0.8171	0.9717	0.9325
IlluminationChanges	0.4628	0.9905	0.0095	0.0372	1.3139	0.4748	0.4673
IntermittentMotion	0.8999	0.9973	0.0027	0.1001	0.6916	0.9543	0.9236
OutOfRange	0.8532	0.9967	0.0033	0.1468	1.0470	0.9110	0.8806
Shadows	0.9290	0.9979	0.0021	0.0710	0.6788	0.9794	0.9515
AVG.	0.8211	0.9955	0.0045	0.1075	1.0854	0.8795	0.8477