Invasive species detection from high resolution aerial photographs

This project focuses on the automatic detection of invasive plant species along railway tracks. Initially, the project targets two highly invasive species: Asian Knotweed (Fallopia japonica) and Giant Hogweed (Heracleum mantegazzianum). ProRail, the Dutch railway infrastructure manager, collects yearly high-resolution helicopter imagery covering all railway tracks in the Netherlands. A deep learning-based classifier is applied to this imagery to detect the aforementioned invasive species. This work was presented as a poster at the 25th anniversary event of the Ambient Intelligence research group.