Application of multi-scale remote sensing imagery to detection and hazard analysis

To both collect terrain data rapidly and save labor costs, the present study employs high-spatiotemporal-resolution imaging through Formosat-2 and aerial photography through unmanned aerial vehicles. In daily visits to the same area, images taken by Formosat-2 can be employed as data for long-term observation. Unclouded images from 2006 to 2010 processed with a false-color overlay and calculated of the normalized difference vegetation index are selected as terrain data. Aerial photos taken by unmanned aerial vehicles are utilized ground-truth data. These two types of data are analyzed to proceed with supervised classification. The results reveal that in the study area mangroves are the most exuberant in summer. The growth of mangroves in Sihcao Wetland remained in dynamic equilibrium from 2006 to 2010 without any sharp increase or decrease in quantity. This proposed method is believed to be suitable to investigate and preserve mangroves by long-term image monitoring and to avoid any unnatural influence on these conservation areas.