Leafsnap is an iPhone app that serves as a mobile field guide for exploring the variety of flora that exists in the northeastern United States. The purpose of this app is to be able to easily identify species of trees and plants simply using visual recognition technology and the camera on an iPhone. The app also contains a fantastically large set of pictures of each of the tree species available in its field guide.
The app comes with a simply enormous set of hi-res photos of leaves, flowers, and fruit. This alone makes the app quite excellent, since it allows the user the opportunity to identify a tree even if he/she does not have internet access. The games component is also quite nice and serves a valuable educational purpose. I am not sure how this can be incorporated into biology classrooms, but it certainly can help students get a better grasp of leaf/flower/fruit identification. The user interface is also incredibly sleek and functional. The final, and most important, piece to this Leafsnap puzzle is the visual recognition technology involved with identifying tree species. The good news is that this feature works. After countless tries, I was able to snap a picture of a leaf on a white background, upload it over wifi, and get an identification.
The key visual recognition feature needs a lot of work. First problem, the need for a strong wifi connection to even get a picture uploaded. The second problem, the need for a white background is a common complaint the app spits back at the user when it claims that it cannot identify the species of the leaf/flower/fruit. It seems like the feature has a lot of potential, and as impressive as it is that this feature even works, it is simply not functional enough at the moment to be effectively used in an educational setting.
This app was created, at least in part, by researchers from Columbia University. There is a lot of room for collaboration here, given our technical chops and the educational nature of this iPhone app. Moreover, I see big opportunity for improvements in their features that rely on visual recognition.
Good app. Has the potential to be a great app. Given that it is for the price of free, I would recommend downloading it and testing it out. I'm sure the app will only get better with time