A hybrid method for accurate iris segmentation on at-a-distance visible-wavelength images

EURASIP Journal on Image and Video Processing

Table 8 Performance comparison with 15 degrees of rotation for each test carried out to assess the robustness

Blur	Noise	Dice	Precision	Recall	Error	Exec. time
0	–	0.9012	0.8853	0.9225	0.0147	1.0580
0	Gaussian	0.8659	0.8327	0.9066	0.0256	0.9668
0	Poisson	0.8876	0.8684	0.9118	0.0221	1.0099
0	Salt & pepper	0.8374	0.8134	0.8848	0.0269	0.9867
3	–	0.8898	0.9073	0.8792	0.0139	1.0714
3	Gaussian	0.8680	0.8538	0.8885	0.0189	0.9406
3	Poisson	0.8859	0.9000	0.8784	0.0162	0.9536
3	Salt & pepper	0.8291	0.8248	0.8549	0.0215	0.9471
6	–	0.7200	0.7815	0.6912	0.0316	0.9543
6	Gaussian	0.7382	0.7514	0.7373	0.0342	0.8789
6	Poisson	0.7114	0.7701	0.6839	0.0289	0.8600
6	Salt & pepper	0.6773	0.7099	0.6859	0.0405	0.8817