Abstract: |
Disaster situations, either natural or man-made, could be catastrophic as causes massive destruction of infrastructures or loss of human lives. First-responders should conduct quickly and efficiently to locate survivors; an operation that becomes difficult when the existing infrastructure and the conventional communication might collapse. Unmanned Aerial Vehicles have emerged as a reliable and cost-effective solution, which aids humans in performing such operations by implementing accurate geographical surveying. The proper combination of hardware and software components (for the design and implementation) has to be selected to produce appropriate results, both with respect to output quality and processing time. In this paper, photogrammetric approaches have been investigated in the terms of collection, processing and producing of dense clouds, orthophotos and digital elevation models considering the aforementioned aspects in disaster risk management missions. Thus, different datasets have been collected where the drone-based flight along with sensing parameters were jointly investigated, evaluating the total processing time and model quality. More specific, the following scenarios were investigated: (i) Selection and combination of the photogrammetric method parameters, (ii) Sensor altitude, and (iii) Area division into separated sectors. The decision about the processing software has be done after assessing the capabilities and the limitations of each solution, since this will affect time, quality and cost of the end result. Under this perspective, a commercial and non-commercial tools compared for the data analysis. Finally, the results of each tool were evaluated while the capabilities and limitations have been perceived. |