Remote Sensing Information Extraction of Impervious Surfaces in Urban Areas Based on Sentinel-2 Data
Abstract：This article uses the Enhanced Normalized Difference Impervious Surface Index (ENDISI) and the Normalized Difference Built-up Index (NDBI) as tools to study the possibility of identifying impervious surfaces from Sentinel-2 satellite remote sensing images. Urumqi, Yining, and Kashgar were selected as study areas, with Sentinel-2 data as the data source. Using remote sensing technology, the change intensity of impervious surface types in urban areas was analyzed from 2020 to 2023. The results showed that after accuracy analysis, the Kappa coefficients of ENDISI were all greater than 0.7, and the extraction accuracy of impervious surfaces in the three cities was higher than 85%; the Kappa coefficients of NDBI were all greater than 0.7, and the extraction accuracy of impervious surfaces in the three cities was higher than 80%. Through comparative analysis with the Chinese 10 m land cover classification dataset for the same time period, it was found that NDBI had a greater distortion degree in the classification results when extracting impervious surfaces. Therefore, ENDISI can more effectively identify impervious surfaces in arid cities.
Keywords：Impervious surfaceEnhanced Normalized Difference Impervious Surface IndexNormalized Difference Built-up IndexSentinel-2
Publication Date：2025-02-27
Pages：5( 1-5 )