Université de Montréal
M.Sc. candidate
Supervisor: Jérôme Théau
François Girard, Université de Montréal
Start: 2024-09-03
Personal page

Project

Detection and mapping of glossy buckthorn (Frangula alnus Mill.) using drone-acquired multispectral imagery and lidar data
Glossy buckthorn (Frangula alnus Mill.) is one of the invasive alien species identified by the Canadian Food Inspection Agency in 2008. It is a shrub known to aggressively colonize the understory following human intervention, such as forest harvesting. Despite the threat it poses to natural recruitment and succession in forest environments and the potential benefits of silvicultural investments, there is still no complete picture of glossy buckthorn distribution in Canada. For my research project, I'm exploring how the fusion of drone-acquired multispectral imagery and lidar data can allow to exploit phenological and structural characteristics simultaneously to obtain a reliable, cartographic portrait of glossy buckthorn in forest environments, particularly in softwood undergrowth. Among other things, I will test an object-based image analysis approach and a convolutional neural network approach to develop an innovative detection and classification method.

Keywords

nerprun, obia, réseau de neurones, fusion de données