Communications of the ACM
Engineers Teach Machines to Recognize Tree Species
Engineers at the Calfornia Institute of Technology have developed a new way to survey urban trees.
Credit: Caltech
A new method developed by engineers from the California Institute of Technology (Caltech) has the potential to change the way urban forests are surveyed.
Caltech professor Pietro Perona says the approach relies on data from satellite and street-level images, such as those from Google Maps, and can automatically create an inventory of street trees.
Working in Pasadena, CA, computer-vision specialists first developed a way to automatically "look" at any specific location in the city using aerial and street-level images from Google Maps, and then created an algorithm to detect objects within these images and calculate their geographic location. Perona says the team used artificial neural networks to train the algorithm to determine which objects were trees.
The group then trained the algorithm to identify 18 of the more than 200 species of trees in Pasadena. The researchers compared the algorithm's results with those of a 2013 tree survey and found the algorithm could detect the identity of a tree's species from Google Maps images with about 80% accuracy.
Perona says cities eventually could use the computer-vision software as part of a long-term technological solution for the management of urban forests.
From Now@Caltech
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found