With rapid urban growth in recent years, the understanding of urban biophysical composition and dynamics has become an important research topic. Remote sensing technologies constitute a potentially scientific basis for examining urban composition and monitoring its changes over the time. The vegetation-impervious surface-soil-water (V-I-S-W) model, in particular, provides a foundation for describing urban/suburban environments and also serves as a basis for further urban analyses comprising urban growth modeling, environmental impact analysis, and socioeconomic factor estimation. This paper developed a normalized spectral mixture analysis (NSMA) method for examining urban composition in Haizhu district using Landsat ETM+  data. In particular, a brightness normalization method was applied to reduce brightness variation. Through this normalization, brightness variability within each V-I-S-W component was reduced or eliminated, thus allowing a single end-member to represent each component. Furthermore, with the normalized image, four end-members, namely vegetation, impervious surface, soil, and water, were chosen to model heterogeneous urban composition using a constrained spectral mixture analysis (SMA) model. The accuracy of impervious surface estimation was assessed and compared with the other existing models. The results indicate that the proposed model is a better alternative to existing models, with a root mean square error (RMSE) of 12.6% for impervious surface estimation in the study area.