Day 4

Detailed paper information

Back to list

Paper title Drone image processing for water quality in the cloud
Authors
  1. Liesbeth De Keukelaere VITO Speaker
  2. Robrecht Moelans VITO
  3. Els Knaeps VITO
Form of presentation Poster
Topics
  • A7. Hydrology and Water Cycle
    • A7.06 EO for monitoring water quality and ecological status in inland waters
Abstract text The use of drones to monitor water quality in inland, coastal and transitional waters is relatively new. The technology can be seen as complementary to satellite and in-situ observations. While cloud cover, low revisit times or insufficient spatial resolution can introduce gaps in satellite-based water monitoring programs, airborne drones can fly under clouds at preferred times capturing data at cm-resolution. In combination with in-situ sampling, drones provide the broader spatial context and can collected information in hard-to-reach areas.

Although drones and lightweight cameras are readily available, deriving water quality parameters is not so straightforward. It requires knowledge of the water optical properties, the atmospheric contribution and special approaches for georeferencing of the drone images. Compared to land applications, the dynamic behavior of water bodies excludes the presence of fixed reference points, useful for stitching and mosaicking and the images are sensitive to sun glint contamination. We present a cloud-based environment, MAPEO-water, to deal with the complexity of water surfaces and retrieve quantitative information on the water turbidity, the chlorophyll content and the presence of marine litter/marine plastics.

MAPEO-water supports already a number of camera types and allows the drone operator to upload the images in the cloud. MAPEO-water also offers a protocol to perform the drone flights and allow efficient processing of the images from raw digital numbers into physically meaningful values. Processing of the drone images includes direct georeferencing, radiometric calibration and removal of the atmospheric contribution. Final water quality parameters can be downloaded through the same cloud platform. Water turbidity and chlorophyll retrieval are based on spectral approaches utilizing information in the visible and Near Infrared wavelength ranges. Drone data are complementary to both satellite and in-situ data. Marine litter detection combines spectral approaches and Artificial Intelligence. Showcases including satellite, drone and in-situ observations will demonstrate the complementary of all three techniques.