Fractal characteristics of very high resolution satellite imagery

Abstract

Differential Box Counting (DBC) model, which is more accurate and can cover the full dynamic range of fractal dimension was employed to measure fractal dimension of image intensity surfaces in this research. The fractal characteristics of panchromatic and multi-spectral image bands of Very High Resolution satellite imagery were analyzed and assessed according to the calculated fractal dimension of typical land cover types including residential area, water body, bare land, woodland, shrubbery, cropland, and grassland in the image. Further discussions on the fractal characteristics of VHR satellite imagery were given by the comparison and analysis with high spatial resolution and medium spatial resolution imagery of the same test site and with the similar acquisition date from the view of multi-scale space and multi-spectral space respectively.