This project done for a graduate project I. System will take a snapshot from plant and will say name of the plant with at least 60% accuracy. I train the system using deep learning architectures(Alexnet, GoogleNet and CaffeNet) for about 1000 kinds of plants. The network trained with ~600k pictures and interface developed with python django frameworks as an interface to the project.
Project demo about how it is working (in Turkish): https://www.youtube.com/watch?v=gpUpMg34T3s