Improving Urban Impervious Surfaces Mapping through Integrating Statistical Methods and Spectral Mixture Analysis

Impervious surfaces have been widely considered as the key indicator for evaluating urbanization and environmental quality. As one of most applied methods, spectral mixture analysis (SMA) has commonly used mapping urban impervious surface fractions. When implementing SMA, original multispectral remote-sensing reflectance images are served foundation to successful SMA. However, limited variances among different land covers from make it challenging in information extraction results unsatisfactory results. To address this issue, a new method proposed study improve through integrating statistical methods In particular, two traditional principal component (PCA) minimum noise fraction rotation (MNF) were highlight covers. Three endmember classes (impervious surface, soil, vegetation) corresponding spectra identified extracted vertices 2-D space plots generated by first three components each PCA MNF. A dataset was stacking MNF (in total six components), fully constrained linear SMA implemented map fractional surfaces. Results indicate that promising performance achieved with systematic error (SE) ?3.45% mean absolute (MAE) 11.52%. Comparative also show much better method-based than conventional