Invasive plants in our lakes!
As a result of nutrient runoff and the introduction of foreign species, some fast-growing plants threaten the lakes and rivers of southern Quebec by modifying habitats, nutrient balance and oxygen levels of aquatic ecosystems. The monitoring of development of invasive species in our lakes is long and costly, delaying the early response needed to prevent further degradation.
The current project proposes the development of aquatic drones adapted for the monitoring of the spatial distribution of underwater herbaceous species. This development will increase our ability to map and monitor the progression of invasive species and curtail the invasion.
A double-shelled self-navigating boat, equipped with a DLSR camera, high-precision RTK positioning system, state-of-the-art navigation units, 3D printed servo to motor interfaces, micro-computer and remote control units, is acquiring the thousands of pictures needed to use Machine Learning to be able to recognize aquatic plants and create mapped plants inventories by itself, during the summer nights.
A software will be developed to facilitate the planning and execution of field missions. A webapp is used to facilitate the classification of the numerous pictures. You can help in creating this useful new aquatic plants dataset.
This innovative approach will improve the management of our lakes, and lead to the publication of articles in biology, computer science and hydrology.
The team will participate in the ongoing design, testing and deployment of the Ubergaiter, the autonomous aquatic drone equipped with submersible underwater camera to identify the species and percentage of surface coverage by each species.
Algorithms will be developed using neural networks for image analysis to identify herbaceous species and produce occupation area maps at 1m resolution.