Assessment of Open-Source Software, QGIS, to Estimate Hurricane Matthew Flood Extent in Robeson County, North Carolina, Using Unsupervised Classification

摘要

The spatial extent of flooding caused by Hurricane Matthew in Robeson County, NC, in October 2016 was investigated by comparing two Landsat-8 images (one flood and one non-flood) following K-means unsupervised classification for each in both ENVI, a proprietary software, and QGIS with Orfeo Toolbox, a free and open-source software. In this study, unsupervised classification was capable of rapidly producing regional maps, but poor accuracy constrained practical application. Of particular note, the open-source setup performed on par with the proprietary option for each of the classifications. Overall, remote sensing techniques using opensource software show promise in helping aid workers to cost-effectively conduct post-event analyses and relief efforts.