Forest Mapper is an iOS smartphone app that attempts to identify trees from images of the trunk, and stores the identification alongside the location to enable the efficient mapping of forests. The identification comes from a convolutional neural network trained on crowd-sourced images of tree trunks. Some phones — those with appropriate depth sensing — also estimate the diameter of the tree, which in turn can be used to estimate its age.
Project team: Gareth Russell, Rajeev Botadra, Pooja Kittanakere Balaji, Smriti Poovannan and Chanales Flores, and the 4th grade students at Bear Tavern Elementary in Hopewell Township, New Jersey!
Current Status: In development. The NJIT team are working on the app mechanics, while the Bear Tavern students are collecting images of tree bark to train the network. We are aiming for 100 images of each of 24 common tree species. You can see their progress on the leaderboard.
Forest Mapper has its own set of pages on this site: check the top menu.