Rubble Volume’s Estimate in Urban Damaged Areas Based on Photogrammetry by Drones

Abstract

Photogrammetry by UAV or drones is a modern, low-cost technology used in several engineering fields. In the current study, this technique was been tested in the field of modeling damaged urban fabric and volume assessment of rubble’s piles. Point cloud is one of the important digital photogrammetric outputs. The segmentation of the main point cloud to sub-clouds representing in a fully carried out automatically, and then estimating the volumes of the stacks stuck under its surface.

The study assumes that point cloud is adopted as a matrix of three-dimensional points without the presence of ancillary data. Segmentation was carried out according to successive phases starting from the filtering of the isolated points, then calculating some necessary amounts of treatment such as the normal’s direction, the local degree of curvature, and the scheme of moving from one point to its vicinity. A proposed customized version of the general method called the growing region is presented; it adopts criteria imposed by morphological characteristics of the rubble piles. The suggested method was applied to a sample of aerial survey data for a destroyed area at the east of Damascus city, and gave accurate results at a minimal cost.