EURASIP Journal on Image and Video Processing

Table 3 F1 results on the ChangeDetection.net dataset dynamic background image sequences

From: Segmentation of moving objects in image sequence based on perceptual similarity of local texture and photometric features

Sequence	Our method	ViBe	SILTP	MC-SILTP
boats	0.566	0.178	0.015	0.127
canoe	0.850	0.783	0.107	0.695
fountain01	0.196	0.061	0.008	0.050
fountain02	0.632	0.563	0.024	0.098
overpass	0.704	0.685	0.066	0.209
Average	0.590	0.454	0.044	0.236
Variance	0.059	0.101	0.002	0.069

The best results are shown in italics

Back to article page