Study on Semantic Contrast Evaluation Based on Vector and Raster Data Patch Generalization

Table 1

Patch number, area change and semantic computation before and after generalization.

	Patch number			Patch area			Semantic consistency		Semantic completeness
	Original; data	Vector; data	Raster; data	Original data; (hm^2)	Vector data; (hm^2)	Raster data; (hm^2)	Vector; data	Raster; data	Vector; data	Raster; data
Cropland	545	258	44	703.797	743.035	718.999	0.944	0.978	0.473	0.081
Garden land	434	176	62	629.161	629.257	644.396	1	0.976	0.406	0.143
Forest land	154	56	21	824.501	823.27	826.414	0.999	0.998	0.364	0.136
Other agricultural lands	41	11	8	41.711	46.554	28.382	0.884	0.68	0.268	0.195
Urban land	242	94	28	234.213	241.244	235.752	0.97	0.993	0.388	0.116
Rural residential land	171	86	41	430.688	444.607	462.983	0.968	0.925	0.503	0.24
Mining land	9	6	5	31.316	31.335	32.728	0.999	0.955	0.667	0.556
Traffic and water conservancy land	37	17	8	134.213	91.865	107.572	0.684	0.801	0.459	0.216
Other construction lands	6	3	3	40.743	40.415	40.181	0.992	0.986	0.5	0.5
Water	1	1	1	1.752	3.33	3.386	0.099	0.067	1	1
Natural reserve land	165	60	35	195.697	172.88	165.305	0.883	0.845	0.364	0.212
Total	1805	768	256	3267.792	3267.792	3266.096	1	0.999	0.425	0.142

Note: all vector data and raster data are the statistical result of generalization of land use patch.