TreeLogy

“Treelogy” is a mobile application, which can perform leaf-based tree identification amongtree species in Turkey using one picture of a given leaf.There is only handful of applications interested in tree identification and they are designedprimarily for detecting North American and European tree species.
There is no applicationwhich has a good performance and localization support for identifying tree species native toTurkey.
This project is aimed to fill this gap.While we are constructing this application, we worked on supervised learning for classi-fication task.
We focused on both Deep Learning (specifically Deep Convolutional NeuralNetworks) and Support Vector Machines.
Tree identification process uses leaf image featuresgathered from Caffe, a convolutional neural network framework, and our image processingmodule.After several experiments, we reached the optimal classification accuracy of 93.59% for 57tree species.
Experiments involve 16096 training and 3020 testing leaf images.
According toour findings, we come to the following conclusion.
Certain image processing procedures forextracting features such as shape and texture descriptors, which we have used in our project,does not produce features as feasible as convolutional neural networks."Created by group paY inekereG"Server update, bug fix