Green stormwater infrastructure (GSI) including bioswales, rain gardens, and wetlands, enhances water quality and alleviates the strain on urban wastewater treatment facilities. However, invasive plants alter the hydrology of the GSI, potentially leading to changes in water infiltration rates and soil erosion, as well as increasing the maintenance costs for removal. Remote sensing using Unmanned Aerial Vehicles (UAVs) and satellites is one approach for quickly and effectively assessing and identifying invasive plant abundance through image classification. This study explores the feasibility of applying UAV imagery ( < 5 cm) and high-resolution satellite imagery ( <30 cm) with machine learning algorithms to classify and identify invasive plants in GSI within Milwaukee, WI. The calibration of classification models using satellites is performed using observations from field surveys and UAV data. Several remote sensing detection techniques will be presented, including spectral detection, texture, and object-based detection for invasive plant species mapping. Ultimately, this study highlights the potential for remote sensing methods in the detection of invasive plant species in GSI for more effective maintenance actions.
