Abstract
Cities are evolving and districts are changing their characteristics faster than ever before. Although the evolution is slow in the central parts of most cities, it is typically fairly fast in outlying regions. They affect the public and private utility networks and maps become less reliable. As a result, emergency plans based on these maps may be ineffective. To assist experts, planners, policy makers, and civil defense organizations, we are developing automated techniques. In previous work, we considered discriminating rural and urban regions [Classifying land development in high resolution satellite images using straight line statistics]. To automate the fine classification process, this paper introduces graph theoretical measures over grayscale images. These measures are monotonic with increasing structure (organization) in the image. Thus, increased cultural activity and land development are indicated by increases in these measures - without explicit extraction of road networks, buildings, residences etc. We present a theoretical basis for the measures followed by extensive experimental results. We consider commercial IKONOS data, which are metric images. Our dataset is large and diverse, including sea and coastline, rural, forest, residential, industrial, and urban areas. On this data set we obtained promising results.