New paper on coastal habitat mapping and machine learning published, soon to be presented at ISPRS Congress
This week, members of the SeaBee team published a paper on coastal habitat mapping and classification and machine learning. The authors include WP3 members Izzie Yi Liu and James Edward Sample, SeaBee Project Lead Kasper Hancke, and SeaBee WP 3 lead Arnt-Børre Salberg. Izzie Yi Liu, Senior Research Scientist at the Norwegian Computing Center and responsible for data and image analysis in the SeaBee project, will be presenting the paper at the International Society for Photogrammetry and Remote Sensing Congress in Nice taking place next week.
The team behind the paper made a unified re-implementation of three neural network models. They then applied distinctive architecture and complexity for general use with Unmanned Aerial Vehicles’ (UAV) multi-sensor data. They tested this on a high-resolution dataset acquired for coastal habitat monitoring and used conventional RGB data and the NIR band from multispectral sensors.
“Society can benefit from this study as by using the trained AI models, highly efficient coastal habitat mapping can be achieved,” said Liu. “It not only will save the marine biology researchers a tremendous amount of manual annotation work, but more significantly, it makes large scale, or even national, coastal habitat mapping possible.”
This work is an important contribution to the SeaBee project, and will help society be better informed with knowledge about our valuable coastal environment and natural resources.