Do you want to identify a plant? The best thing you can do right now is take a photograph and search for similar results on Google, but if facial recognition has achieved such precision these days, why couldn't we do the same with the world of flora? That concept brings us to Pl@ntNet, a mobile app that compares your captures against more than twenty-one thousand species in a giant, collaborative, and constantly expanding database.
Some people are said to have a “green thumb” for plants... and I am definitely not one of them. While I need my dose of nature like any other human being, the truth is that I can probably recognize two or three types of flowers at most, not counting of course the vegetables that usually end up in a salad. When in doubt, the most effective resource for the user is to take a picture of the flower in question, share it on some social network hoping a user with some botanical experience will identify it, or make a direct comparison on Google Images. However, there is another option: Pl@ntNet.
How to Identify Plants with a Photograph
Pl@ntNet is an app compatible with iOS and Android (and with a web version) that allows us to load the image of a plant into a special database and search for any available match. If there is no match, the photo waits for another user to identify the species.
Users have the ability to rate both images and the identification made, and the highest quality results appear first in the app. Unlike other solutions based on algorithms and artificial intelligence, Pl@ntNet works collaboratively.
Some features of the platform are:
- The ability to filter recognized species by genus or family
- The app gives more weight to the opinion of users as they prove their knowledge
- Multi-flora identification to search for the plant in all types of flora
- Browsing the galleries using the taxonomic category
To test it, we used this photograph taken from the patio of one of our collaborators and submitted it to the tool's analysis, in its web version. The results, which you can see in the photo above, were obtained quickly and with astonishing precision. Definitely, the tool works.
As you will see, the potential of this app is truly impressive, as it not only can assist enthusiasts, but could be of great use in professional settings, such as an identification process at a customs post to prevent the entry of a destructive species into a certain region. Its collaborative engine already has 21,963 species identified with 2,272,213 photographs in the database.
