A robust and consistent stack generalized ensemble-learning framework for image segmentation

Journal of Engineering and Applied Science

Table 3 The performance comparison using different numbers of base learners and the proposed model on test images (best results are highlighted using boldface)

Image	Metrics	LightGBM	SVM	XGBoost	Proposed
12,003	CC	0.0808	− 0.0005	0.0689	0.1146
	SAM	0.43924	0.4383	0.4412	0.45149
	SSIM	89.2904	89.1634	89.2735	89.3078
	UQI	0.2079	0.1939	0.2103	0.2356
106,005	CC	− 0.0591	0.0032	− 0.0583	− 0.0915
	SAM	0.4382	0.4537	0.4382	0.4606
	SSIM	89.3817	89.3745	89.3823	89.3371
	UQI	0.31457	0.2997	0.3141	0.3459
113,044	CC	0.1278	0.0040	0.2279	0.3979
	SAM	0.5950	0.4125	0.4497	0.6925
	SSIM	89.6365	89.5298	89.6325	89.3997
	UQI	0.5901	0.3309	0.3679	0.6094
107,014	CC	0.0201	0.0062	0.2704	0.2760
	SAM	0.4213	0.3982	0.4297	0.4518
	SSIM	89.6396	89.5689	89.6244	89.6115
	UQI	0.2994	0.2867	0.3125	0.3335