Flavia, A Leaf Recognition Algorithm/ Program for Plant Classification using PNN (Probabilistic Neural Network)

Flavia

a Leaf Recognition Algorithm/ Program for Plant Classification using PNN (Probabilistic Neural Network)

This webpage is under construction. Documents and the program itself are of version alpha. We will do our best to make them better. But due to we all have many research work to do, this part-time project will progress slowly.

Introduction:

The purpose of this program is to teach a computer to classify plants via their leaves. You just need to input the leaf image of plant(acquired via digital camera or scanners), then the computer can tell you what kind of plant it is.

We utilize the PNN (Probabilistic Neural Network) to implement this AI process. 12 characters of leaves are taken into account, including geometrical ones and morphological ones. After discriminant analysis (stepwise method), all these characters are reserved. PCA orthogonalizes these 12 characters into 5 principal variables, which are input vectors of the PNN.

Presently, our system can classify 32 common plants in Yangtze Delta(Shanghai is just there), China: Phyllestachys Pubescens, Aesculus Chinensis, Berberis Ferdinandi-coburgii Schneid, Cercis Chinensis, Indigofera Tinctoria Linn., Acer Dalmatum, Phoebe Zhennan, Kalopanax Septemlobus, Cinnamomum Japonicum, Koelreuteria Paniculata, Ilex Macrocarpa, Pittosporum Tobira, Chimonanthus Praecox, Cinnamomum camphora, Viburnum Awabuki, Osmanthus Fragrans, Cedrus Deodara, Ginkgo Biloba, Lagerstroemia Indica, Nerium Indcus, Podocarpus Macrophyllus, Prunus Yedoensis, Ligustrum Lucidum Ait., Toona Sinensis, Amygdalus Persica, Manglietia Fordiana, Acer Buergerianum, Mahonia Bealei, Magnolia Grandiflora, Populus Xcanadensis Moench, Liriodendron Chinense, Citrus Reticulata Blanco. The average accuracy is 93% for all of them.

The program is written in MATLAB. Thus the program can work on Linux, MacOS, Solaris and Windows. We also provide the Linux binary executables to x86 Linux users.

Download. Read docs first

Before you download our software, be sure to read this document for download instructions first. It describes the file structure, download options, required libraries, etc. Since there is many components of our program, the docs will not mess you up. It is also the user manual of this program. If you go to the download page, you will see a table like this.

Please download the latest document file which is named as Flavia-doc-< version > You can choose to download the odt file(can be opened by OpenOffice 2 or later) or PDF file.

This link will direct you to the download page.

The latest document is Flavia-doc-alpha.odt or Flavia-doc-alpha.pdf. Read it first.

Copyleft

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. http://www.gnu.org/copyleft/gpl.html