Signal Processing and Speech Communication Laboratory
hometheses & projects › Stockpile Volume Estimation using Photogrammetry

Stockpile Volume Estimation using Photogrammetry

Status
Finished
Type
Master Thesis
Announcement date
16 Jan 2024
Student
Stefan Fragner
Mentors
Research Areas
 Getting a somewhat accurate number for the volume of a possibly large stockpile of material can be difficult. Usually this has to be done using specialized equipment like 3D laser scanners. This kind of equipment is expensive, and it can be time-consuming to set up the measurement process. Something that is however inexpensive and easy to use is a smartphone. Today almost every person carries a smartphone and many companies even supply their employees with them. Modern smartphones can take pictures at high resolutions which can be used to compute accurate 3D models of photographed objects. These models can be used to compute various metrics such as the volume of a stockpile. In this work, a software was implemented to perform exactly this kind of computation. The developed software uses photogrammetry together with circular coded targets to reconstruct an accurate and scaled 3D model from a set of images. This 3D model is segmented to identify the stockpile and determine its volume. Multiple experiments were performed, and it was shown that the estimated volume has an error of about less than 3%. The software also provides a graphical user interface for easy execution of all functionality.