Improving Minimum Cross-Entropy Thresholding for Segmentation of Infected Foregrounds in Medical Images Based on Mean Filters Approaches

Table 4

Comparison of the unsupervised and supervised evaluation for the segmented images in Figure 8.

	Unsupervised evaluation			Supervised evaluation
	RC	IU	IRD	JI	FS	QLTY
Img.1(c)	0.8709	0.8782	0.8462	—	—	—
Img.1(d)	0.8816	0.8852	0.8398	—	—	—
Img.1(e)	0.8897	0.8989	0.8837	—	—	—
Img.2(c)	0.8630	0.8702	0.8529	—	—	—
Img.2(d)	0.8759	0.8778	0.8408	—	—	—
Img.2(e)	0.8879	0.8899	0.8873	—	—	—
Img.3(c)	0.8519	0.8458	0.8381	0.8379	0.9063	0.9173
Img.3(d)	0.8815	0.8689	0.8473	0.8449	0.9037	0.9290
Img.3(e)	0.8857	0.8939	0.8685	0.8559	0.9131	0.9792
Img.4(c)	0.8498	0.8500	0.8482	0.8398	0.8971	0.9157
Img.4(d)	0.8698	0.8697	0.8688	0.8407	0.9026	0.9238
Img.4(e)	0.8878	0.8941	0.8705	0.8518	0.9099	0.9672
Img.5(c)	0.7509	0.7558	0.7238	0.7210	0.8308	0.7294
Img.5(d)	0.8647	0.8794	0.8600	0.9325	0.9227	0.9371
Img.5(e)	0.8788	0.8899	0.8728	0.9497	0.9549	0.9503
Img.6(c)	0.7618	0.7592	0.7327	0.7292	0.8299	0.7381
Img.6(d)	0.8797	0.8704	0.8599	0.9293	0.9208	0.9305
Img.6(e)	0.8828	0.8878	0.8779	0.9431	0.9459	0.9492