2025_programme: NautiloPy: An Open-Source Python framework for the Gironda Underwater Cave Sonar And Vision Data set

Back


  • Day: June 17, Tuesday
      Location / Time: B. ERATO at 17:00-17:20
  • Last minutes changes: -
  • Session: 28. Signal and Image Processing
    Organiser(s): N/A
    Chairperson(s): Gabriele Morra
  • Lecture: NautiloPy: An Open-Source Python framework for the Gironda Underwater Cave Sonar And Vision Data set
    Paper ID: 2185
    Author(s): Thomas Guilment, Gabriele Morra, Leonardo Macelloni, Marco D'Emidio
    Presenter: Gabriele Morra
    Abstract: Ocean exploration research, particularly in complex environments like underwater caves, faces challenges, including the complexities of exploration and mapping, limited publicly available datasets, and a lack of tools for processing and visualizing existing data. These obstacles often result in experienced researchers spending a disproportionate amount of time on data preparation alone before meaningful analysis can begin. For newcomers or interdisciplinary researchers, technical barriers can even render existing data unusable. To tackle this challenge, we created NautiloPy, an open-source Python project designed to democratize underwater acoustic data processing and visualization.\n\nNautiloPy's capabilities are demonstrated using the Gironda Underwater Cave Sonar and Vision Data Set (Mallios et al., 2017), a real-world benchmark dataset featuring Autonomous Underwater Vehicle (AUV) navigation data, sonar measurements, and video recordings. The project offers well-written codes, documentation, and notebooks for data pre-processing (e.g., 3D AUV pose, orientation interpolation, and sonar cleaning), interactive visualization (e.g., optimized cached animation of sonar scanning over time) from the raw data to the final 3D map of the Gironda underwater cave. NautiloPy is easy to set up locally using the UV Python manager from Astral. It is also available online on Google Collaboratory, allowing users to run the project without any setup.\n\nBy bridging the gap between raw data and accessible tools, NautiloPy aims to establish a new standard for open-access underwater data processing while inspiring educational outreach and interdisciplinary research such as automatic sonar data cleaning, combining sonar data with underwater photogrammetry for better 3D reconstruction, 3D map interpolation technics and advanced GPU computing for 3D visualization, to cite a few.
  • Corresponding author: Dr Thomas Guilment
    Affiliation: University of Louisiana at Lafayette
    Country: United States

Back