Revealing Intra-Urban Features using Optical and SAR Images

Abstract

We developed a method for revealing the intra-urban features and estimating population density using remotely sensed imagery. Remotely sensed imagery has been widely used in urban studies. However, the moderate-resolution optical data (e.g. Landsat) is often too coarse for delineating urban features, since urban areas are far more heterogeneous than most other land cover types. Therefore, we employed both optical and synthetic aperture radar images for estimating population density inside urban areas. This research indicated that the integrated features of the optical and SAR images can increase the accuracy for estimating population density.