A Neural Network-Based Approach to Multiple Wheat Disease Recognition

Igor V. Arinichev; Sergey V. Polyanskikh; Irina V. Arinicheva; Galina V. Volkova; Irina P. Matveeva

doi:10.5391/IJFIS.2022.22.1.106

Original Article

Split Viewer

International Journal of Fuzzy Logic and Intelligent Systems 2022; 22(1): 106-115

Published online March 25, 2022

https://doi.org/10.5391/IJFIS.2022.22.1.106

A Neural Network-Based Approach to Multiple Wheat Disease Recognition

Igor V. Arinichev¹, Sergey V. Polyanskikh², Irina V. Arinicheva³, Galina V. Volkova⁴, and Irina P. Matveeva⁴

¹Department of Theoretical Economy, Kuban State University, Krasnodar, Russia
²Plarium Inc., Krasnodar, Russia
³Department of Higher Mathematics, Kuban State Agrarian University named after I.T. Trubilin, Krasnodar, Russia
⁴Laboratory of Cereal Crops Immunity to Fungal Diseases, All-Russian Research Institute of Biological Plant Protection, Krasnodar, Russia

Correspondence to :
Igor V. Arinichev (iarinichev@gmail.com)

Received: November 17, 2021; Revised: January 13, 2022; Accepted: February 4, 2022

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Go to

Abstract
Acknowledgments
Conflict of Interest
Biographies

In this paper, modern computer vision methods are proposed for detecting multiple diseases in wheat leaves. The authors demonstrate that modern neural network architectures are capable of qualitatively detecting and classifying diseases, such as yellow spots, yellow rust, and brown rust, even in cases in which multiple diseases are simultaneously present on the plant. For certain classes of diseases, the main multilabel metrics (accuracy, micro-/macro-precision, recall, and F1-score) range from 0.95 to 0.99. This indicates the possibility of recognizing several diseases on a leaf with an accuracy equal to that of an expert phytopathologist. The architecture of the neural network used in this case is lightweight, which makes it possible to use offline on mobile devices.

Keywords: CNN, Multilabel classification, Wheat diseases, Computer vision

Acknowledgments

Go to

Abstract
Acknowledgments
Conflict of Interest
Biographies

The research was carried out with the financial support of the Kuban Science Foundation under the framework of scientific project (No. IFR-20.1/75).

Conflict of Interest

Go to

Abstract
Acknowledgments
Conflict of Interest
Biographies

No potential conflict of interest relevant to this article was reported.

Biographies

Go to

Abstract
Acknowledgments
Conflict of Interest
Biographies

Igor V. Arinichev is a candidate of Economic Sciences, associate professor of the Kuban State University. He published more than 60 publications in peer-reviewed journals or conferences, and books on application of mathematical methods in economics, agriculture and technology.

E-mail: iarinichev@gmail.com

Sergey V. Polyanskikh is a Ph.D. in Mathematics and Mechanics. He is currently a senior data scientist in Plarium Inc. He published more than 20 theoretical and applied publications in hydrodynamics and mathematics.

E-mail: spmathf@gmail.com

Irina V. Arinicheva is a doctor of Biological Sciences, and a professor of the Department of Higher Mathematics (Kuban State Agrarian University). Her specialization is mathematical modeling of biological processes. She is the author of over 150 scientific articles, monographs, inventions, educational materials for students.

E-mail: loukianova7@mail.ru

Galina V. Volkova is a doctor of Biological Science, and Head of the Laboratory of Immunity of Cereal Crops to Fungal Diseases. She published more than 200 publications in peer-reviewed journals or conferences, and books.

E-mail: galvol2011@yandex.ru

Irina P. Matveeva is a researcher of the Laboratory of Immunity of Cereal Crops to Fungal Diseases, postgraduate student, and a author of over 10 publications.

E-mail: irina.matveeva14@yandex.ru