A framework for retinal vasculature segmentation based on matched filters

Table 3 Performance results on DRIVE database images

Image	Sen	Spe	PPV	NPV	ACC	Kappa
1	0.7928	0.9770	0.8359	0.9697	0.9534	0.7871
2	0.7646	0.9893	0.9254	0.9604	0.9561	0.8123
3	0.6927	0.9859	0.8802	0.9555	0.9478	0.7462
4	0.7661	0.9830	0.8677	0.9665	0.9555	0.7886
5	0.7739	0.9854	0.8766	0.9703	0.9605	0.7999
6	0.6939	0.9853	0.8805	0.9537	0.9459	0.7459
7	0.7972	0.9764	0.7967	0.9765	0.9578	0.7734
8	0.7285	0.9732	0.7461	0.9707	0.9493	0.7092
9	0.7343	0.9791	0.8234	0.9653	0.9505	0.7486
10	0.7737	0.9823	0.8357	0.9739	0.9606	0.7816
11	0.7671	0.9839	0.8670	0.9686	0.9578	0.7903
12	0.7614	0.9806	0.8378	0.9689	0.9551	0.7726
13	0.6780	0.9858	0.8911	0.9468	0.9406	0.7367
14	0.7846	0.9757	0.7985	0.9737	0.9549	0.7662
15	0.7318	0.9768	0.7926	0.9677	0.9502	0.7333
16	0.7771	0.9807	0.8501	0.9690	0.9556	0.7869
17	0.8041	0.9736	0.7864	0.9762	0.9553	0.7700
18	0.7519	0.9868	0.8977	0.9628	0.9555	0.7932
19	0.6961	0.9910	0.9287	0.9509	0.9485	0.7670
20	0.7074	0.9832	0.8674	0.9559	0.9462	0.7490
Average	0.7489	0.9818	0.8493	0.9652	0.9529	0.7679

ISSN: 1475-925X