Estimation of fractional vegetation cover based on spatial continuity within heterogeneous karst environments

The complex landscapes that make up karst environments are characterized by high ratios of exposed bedrock, non-contiguous soil, and fragmental vegetation. This heterogeneity makes accurate estimation of vegetation coverage difficult. Here, an estimation method based on spatial continuity is proposed and implemented in a typical karst landscape. The proposed image segmentation method accounted for spectral similarity and the spatial relationships of adjacent pixels simultaneously. The comparison results showed that vegetation coverage in the spatially heterogeneous karst environments tended to be underestimated by the normalized difference vegetation index (NDVI) and overestimated NDVI-SMA (spectral mixture analysis) model. The proposed method based on spatial continuity within a complex landscape was found to ameliorate the effects of the highly heterogeneous landscape on the extraction of vegetation coverage from hyperspectral imagery.