Research on extraction method of urban impervious surface information

Abstract

Impervious surface is a character of urban areas. Its cover conditions, changing with urban development, have a great impact on the local climate, hydrology, and the surface energy flux of the city. Based on analyzing the theory of the principal component analysis (PCA), built up area index, linear spectral unmixing, and classification and regression tree to extract impervious surface of Tangshan city, Hebei Province, from Landsat OLI_TIRS image. And by using random sampling points combined with the high resolution remote sensing image, the above four extraction methods are used to extract the impervious surface precision evaluation results. The results show that using the method of classification and regression tree model to extract the impervious surface result accuracy of other classification precision is increased by 3% to 10%.