ABSTRACT
Remotely sensed imagery has been used as input data in regional science applications in urban settlements since the late 1950s. Thus far, moderate spatial resolution (10-100 m) imagery provides good spectral resolution with multiple bands in the visible portion of the electromagnetic spectrum, with at least one band located in the infrared portion of the spectrum and a panchromatic band. In this chapter, several algorithms that are widely used for characterization of urban landscape using medium-resolution imagery are introduced. Section 3.2 describes the most commonly used medium-resolution satellite images for characterization of urban land cover. Section 3.3 explains heterogeneous features of urban landscapes. Section 3.4 outlines algorithms to characterize urban land cover with the use of moderate-resolution imagery. Section 3.5 introduces neutral network and support vector machine (SVM) for urban land cover characterization. Section 3.6 is the summary of the chapter.
