Assessment of Difference in Structure from Motion for Stockpile Volume Estimation using UAV Approach

Abstract

The software for processing the photogrammetry software is increasingly capable of producing a high-quality 3D model DEM, DSM, and DTM from an aerial photo and measuring the volume of material. Nowadays, UAV technology has been rapidly used for stockpile volume calculation for the mining and construction industry. The purpose of this study is to compare image processing for the stockpile volume calculation using three different software for the UAV with the conventional method GPS as a reference. The result obtained using GPS is processed using CIVIL3D and act as a base measurement of the stockpile volume to compare result obtain from photogrammetry software which is Pix4D, Agisoft, and SimActive’s Correlator3D. The analysis of this study compared 3D model for volume calculation from a photo image for each photogrammetry software selected and obtained the result volume calculation for photogrammetry with the accuracy achieving below 3% error from the referenced volume of the stockpile. 12 GCP were placed around the study area to test the difference between the numbers of GCP processing for the volume stockpile results. The image acquisition is taken at 10 meter altitude and 20 meter altitude using UAV Phantom 4 pro and set with Pix4Dcapture circular mission. The processing method is SfM in the 3 photogrammetric selected that have 4 types of processing for each software, 10 meter altitude and 20 meter altitude with 4 GCP and 12 GCP. 12 results of the 3D model and volume of the stockpile was obtained. As for the volume results, the Pix4D 10 meter 4 GCP result obtained the closest volume to the reference. The point cloud generated is also stable based on the cloud compare volume density result. Overall, the study can conclude that not only the altitude of flight affected, and the type of software used is affecting the number of point clouds generated but also the number of GCP.