EURASIP Journal on Advances in Signal Processing

Table 3 F -measure and minimal error $δ_{min}$ (given in brackets) obtained using different local propagation costs

From: Self-adaptive algorithm for segmenting skin regions

γ _p ↓		ECU data set			HGR data set
	$γ_{Δ}^{E} ↓$ $γ_{Δ}^{F} \to$	$γ_{Δ}^{(HL)}$	$γ_{Δ}^{(RGB)}$	$γ_{Δ}^{(SP)}$	$γ_{Δ}^{(HL)}$	$γ_{Δ}^{(RGB)}$	$γ_{Δ}^{(SP)}$
Gaussian
RP	$γ_{Δ}^{(HL)}$	0.8511 (8.20%)	0.8547 (7.71%)	0.8460 (8.05%)	0.9132 (6.75%)	0.9130 (6.75%)	0.9157 (6.55%)
	$γ_{Δ}^{(RGB)}$	0.8469 (8.58%)	0.8511 (8.00%)	0.8440 (8.28%)	0.9136 (6.73%)	0.9134 (6.73%)	0.9159 (6.55%)
	$γ_{Δ}^{(SP)}$	0.8499 (8.43%)	0.8535 (7.96%)	0.8419 (8.44%)	0.9136 (6.73%)	0.9135 (6.74%)	0.9155 (6.57%)
DSPF	$γ_{Δ}^{(HL)}$	0.8 4 1 1 (7.2 2 %)	0.8409 (7.10%)	0.8339 (7.77%)	0.9 5 6 2 (2.5 2 %)	0.9538 (2.79%)	0.9545 (2.55%)
	$γ_{Δ}^{(RGB)}$	0.8425 (7.32%)	0.8420 (7.26%)	0.8355 (7.79%)	0.9564 (2.50%)	0.9540 (2.78%)	0.9548 (2.56%)
	$γ_{Δ}^{(SP)}$	0.8388 (7.34%)	0.8387 (7.22%)	0.8360 (8.04%)	0.9566 (2.49%)	0.9541 (2.76%)	0.9301 (5.70%)
Histogram
RP	$γ_{Δ}^{(HL)}$	0.8330 (9.55%)	0.8389 (8.83%)	0.8341 (9.08%)	0.9204 (5.99%)	0.9204 (5.98%)	0.9220 (5.84%)
	$γ_{Δ}^{(RGB)}$	0.8286 (9.71%)	0.8320 (9.14%)	0.8283 (9.39%)	0.9221 (5.84%)	0.9220 (5.84%)	0.9226 (5.78%)
	$γ_{Δ}^{(SP)}$	0.8313 (9.69%)	0.8 3 4 8 (9.2 0 %)	0.8266 (9.60%)	0.9221 (5.78%)	0.9 2 2 0 (5.7 8 %)	0.9215 (5.82%)
DSPF	$γ_{Δ}^{(HL)}$	0.8353 (7.84%)	0.8347 (7.81%)	0.8295 (8.28%)	0.9555 (2.50%)	0.9539 (2.74%)	0.9520 (2.97%)
	$γ_{Δ}^{(RGB)}$	0.8363 (8.01%)	0.8355 (8.01%)	0.8305 (8.35%)	0.9550 (2.56%)	0.9537 (2.79%)	0.9517 (3.04%)
	$γ_{Δ}^{(SP)}$	0.8350 (7.85%)	0.8342 (7.82%)	0.8322 (8.70%)	0.9544 (2.57%)	0.9530 (2.78%)	0.9372 (4.70%)

The arrows indicate the column or row. Italicized values indicate the configuration used in the remaining experiments.

Back to article page