 This study used a novel and state-of-the-art image segmentation machine learning technique, DeepLab, to analyze two years of high-resolution drone-based imagery of a marine flowering plant species, eelgrass, Zostera marina, an estuary that has experienced large eelgrass declines and subsequent recovery over the past decade. The DeepLab model was able to accurately classify eelgrass at different scales, ranging from meadowscale to smaller patches less than a meter in size. The model's accuracy was comparable to or better than other models for similar seagrass systems, providing valuable data for long-term management strategies seeking to protect these vital ecosystems. This article was authored by Krititalam, Namangwin, Jonathan Ventura, and others.