Evaluation of urban sprawl from space using open source technologies

摘要: Abstract Up to nowadays, satellite data have become increasingly available, thus offering a low cost or even free of charge unique tool, with great potential for quantitative assessment urban expansion and sprawl, as well monitoring land use changes soil consumption. This growing observational capacity has also highlighted the need research efforts aimed at exploring offered by processing methods algorithms, in order exploit much possible this invaluable space-based source. The work herein presented concerns an application study on process sprawl conducted ASTER data. selected test site is highly significant, being it coastal zone (with presence sand rocks) characterized fragmented ecosystem small towns, increasing rate urbanization In produce synthetic maps areas, images were classified using two automatic classifiers, Maximum Likelihood (MLC) Support Vector Machines (SVMs) applied changing setting parameters, aim compare their respective performances terms robustness, speed accuracy. All steps been developed integrating Geographical Information System Remote Sensing, adopting open source software. Results pointed out that SVM classifier RBF kernel was generally best choice accuracy higher than 90%) among all configurations compared, multiple bands globally improves classification. One critical elements found case given mixed rocks. different SVMs, i.e. kernels values allowed us calibrate cope specific need, our case, achieve reliable discrimination from area.