ETM+ image classification research based on Stratified and regional classification method - take land-use and land-cover in Shangri-la for example

摘要

Combine the stratified and regional thinking, we proposed a stratified and regional classification method that based on the spatial resolution fusion and NDVI, and utilized in the classification research of land-use and land-cover in Shangrila. The method first used ETM+453 synthesized image, fused the panchromatic band to promote the spatial resolution, calculated NDVI after the extraction of related ground objects, then divided it into vegetation area and non-vegetation area, and classified the region separately with supervised classification. The results showed that this method, compared with the conventional supervised classification, the classification accuracy enhanced 12.94%.