According to a 2019 survey by the French National Union of Landscape Enterprise, nearly 70% of French people believe that there are not enough plants in the city center. However, before planting new trees, we must already know the existing heritage in order to think better about future plantings (species selection, choice of places to plant, etc.). The inventory of trees in the city is a key step in the management of urban plants.
Nevertheless, this inventory is a tedious and often painful step for managers, who are looking for new tools to support them. Automation of the individual tree detection process could save considerable time in urban greening projects.
To meet this need, we propose to build an open-source individual tree detection tool that can be integrated into a larger software solution. Then create algorithms to categorize individual trees detected on different criteria (essence, size, age ...) with good performance especially on data from local trees. Finally, we want to build a collaborative and open contribution guide to enable citizens and/or professionals to add information.
Right now, we're in the training stage of individual detection algorithms, but we're open to any proposal or help you can give us!
If you're interested in this topic, come and talk to us about our Slack space: https://hack4nature.slack.com/archives/C01B4D71469
