Thesis subject

Qualitative assessment of ground surfaces with AI and UAV technology

The understanding of soils mechanics is an important factor to successfully carry out tasks where there’s some sort of interaction with the ground surface or inner layers, e.g., safety, agriculture. Several soil measurements and tests are carrying out on-site or in lab to determinate the soil type. The geologists use several different types of soil test to classify soils according with international classification standards, e.g., Unified Soil Classification System (USCS) and European soil classification system (ISO 14688). These classification systems usually classify soils according with texture, grain size, shapes, mixtures, through specific tests, e.g., Shear test.

This project aims to verify the following hypothesis: Is it possible to estimate soil ground surface using aerial remote sensing technology?

Objectives

  1. Qualitative analysis from the datasets obtained from the aerial surveys and the on-site measurements.
  2. Quantitative analysis from the correlation between ground and aerial measurements.
  3. Implementation of a machine learning approach to estimate surfaces characteristics remotely.

    Literature

    • D. Anthony, E. Basha, J. Ostdiek, J. P. Ore and C. Detweiler, "Surface classification for sensor deployment from UAV landings," 2015 IEEE International Conference on Robotics and Automation (ICRA), Seattle, WA, 2015, pp. 3464-3470.

    Requirements (optional)

    • Willing to learn about machine learning
    • Enthusiast about field work
    • Excited for doing out-of-the-box experiments

      Theme(s): Sensing & measuring