Geotechnical engineers design and analyze the interface between the natural and built environments, often based on uncertain and inherently variable data due to the nature of the building material, soil. Engineers must estimate the subsurface conditions at a site using historical, in-situ, and laboratory testing data. However, only discrete representative...
Drones are being used increasingly in the transportation sector for roadside applications such as surveying, structural inspection, and traffic monitoring, which raises distraction concerns for drivers. This study examines the visual attention of a group of Oregon drivers to determine drones’ potential as distracting roadside elements. Eye tracking data was...
Seamless, high-resolution topobathymetric digital elevation models (DEMs) of rivers are important for generating hydrodynamic models, studying water quality and related ecological issues, modeling flood vulnerability, and numerous other applications. However, such DEMs can be difficult to generate, due to a lack of topographic and bathymetric source data. Even where the...
Marine debris is a chronic and increasingly pervasive problem for coastal regions around the world. Debris poses environmental risks, threats to wildlife, and degradation of the natural environment. Recent research has shown the advantages of uncrewed aircraft systems (UAS) for detection and recognition of marine debris, including the ability to...
Unmanned aircraft systems (UAS), also known as drones, are an increasingly popular method of collecting surveying and mapping data. Two common drone-based mapping techniques are lidar and structure from motion (SfM) photogrammetry, and a frequently-heard (yet nearly impossible to answer) question is: “which is better?” The most common metric for...
A fundamental problem in computer vision is to partition an image into meaningful segments. While image segmentation is required by many applications, the thesis focuses on segmentation of computed tomography (CT) images for analysis and quality control of composite materials. The key research contribution of this thesis is a novel...
Aligning Real-Time Networks (RTNs) with the National Spatial Reference System (NSRS) is crucial to ensure accuracy, consistency, interoperability, and long-term data integrity for geospatial applications. However, the effectiveness of different alignment methods as well as the need for monitoring approaches to maintain alignment consistency with the NSRS have not yet...
Frequent 3D spatial mapping of estuaries is critical to understanding and modeling changes that can impact ecological health (e.g., water quality) and ecosystem services (e.g., aquaculture, nutrient cycling, and shoreline stabilization). However, obtaining accurate, repeat topobathymetric data for estuaries is a major challenge. Because shallow estuaries are typically not used...
Deep learning has recently revolutionized robot perception in many canonical robotic applications, such as autonomous driving. However, a similar transformation has yet to occur in more harsh environments including underwater and underground. This is due in part to the difficulty in deploying robots in these environments, which lack large real...
During the Canterbury Earthquake Sequence (CES) in Christchurch, NZ, an estimated $15B in damages occurred due to liquefaction-induced phenomena, illustrating the importance of understanding site-specific responses to seismic ground motions, as well the importance of accurately predicting liquefaction severity. One potential method for improving the understanding of how a site...