Unmanned Aircraft Systems-Based Photogrammetry for Ground Movement Monitoring

A major concern during installation or rehabilitation of buried pipelines are ground movements during construction. Unexpected ground movements, such as settlement or uplift, may be the result of geotechnical conditions, proximity to other structures and utility lines, contractor methodologies, and/or poor designs. Near-continuous monitoring of ground movements during construction is important to quickly identify and mitigate potential problems. A common method for detecting ground movement is to routinely measure the coordinates of discrete points along the ground surface during construction using high-accuracy survey equipment, such as automatic levels, total stations, and Global Navigation Satellite System (GNSS) receivers. It is then possible to determine if the position of these points are changing with time. Although these survey equipment provide coordinates with millimeter- to centimeter-level accuracies, the equipment and often the surveyor must physically occupy the discrete points along a construction corridor in order to take the measurements. This paper presents a procedure for measuring and monitoring ground changes by utilizing digital photographs taken from an inexpensive, consumer-grade camera mounted on a small unmanned aircraft system (sUAS). The digital photographs were aligned and processed using a computer vision procedure known as Structure-from-Motion (SfM), which results in ortho-rectified, mosaicked aerial images and high-resolution digital elevation models (DEMs) of the construction site. After a series of UAS flights over time, the resulting DEMs from this method were differenced in order to detect and quantify ground movements. In an effort to define accuracy, results of the UAS-based measurements were then compared with measurements taken from a total station survey. The comparison yielded a vertical root-mean-square deviation (RMSD) of 2 cm at the 95% confidence level.

Language

  • English

Media Info

  • Media Type: Web
  • Features: References;
  • Pagination: pp 1000-1011
  • Monograph Title: Pipelines 2016: Out of Sight, Out of Mind, Not Out of Risk

Subject/Index Terms

Filing Info

  • Accession Number: 01608178
  • Record Type: Publication
  • ISBN: 9780784479957
  • Files: TRIS, ASCE
  • Created Date: Jul 14 2016 3:04PM