Improvement of Volume Estimation of Stockpile of Earthworks Using a Concave Hull-Footprint

In the estimation of volume of stockpiles of earthworks, the question is no longer whether the data collected is dense and accurate (equipment and techniques capable of accurate data measurement are available), but how to manipulate the data to yield accurate volume estimation. Although surface modeling through TIN yields more accurate volumes than grid modeling, the delineation of footprint of the stockpile remains one of the main sources of errors in volume determination due to spurious surfaces created within the convex hull of the TIN model. In this paper, an approach for automatic delineation of the stockpile footprint based on a concave hull is introduced. A concave hull as a geometry (usually point data) container is realized by minimizing the enclosing planimetric area and it is usually not unique. Several algorithms for creating concave hulls are suggested, in this paper an algorithm based on Delaunay triangulation and linear referencing was used to create the concave hull. A comparison of volume estimations of stockpiles taking into consideration the footprint via convex hull, concave hull and manually delineated outline showed that volumes based on the concave hull are closer in value to volumes based on manually delineated footprint. Therefore in the absence of points manually picked to represent the outline of a footprint, the concave hull can be relied on.