DetecTree: Tree detection from aerial imagery in Python

Urban tree canopy datasets are important to a wide variety of social and environmental studies in cities. Surveys to collect such information manually are costly and hard to maintain, which has motivated a growing interest in automated approaches to tree detection in recent years. To that end, LIDAR point clouds are a very appropriate data source, especially given its capability to represent spatial features in three dimensions. However, collecting LIDAR data requires expensive equipment and raw datasets are rarely made openly available. On the other hand, aerial imagery is another major source of data that is able to capture a wide range of objects on Earth, including trees. Although aerial imagery depicts the spatial features in only two dimensions, its main advantage with respect to LIDAR is its greater availability.