According to the FAO, around 660 million people might still suffer from chronic hunger in 2030. Our food production must dramatically increase to meet the ever-growing demand of the population. However, plant diseases are a major reason behind our food insecurity, and it is estimated to rise due to globalization, global warming, etc. Thus, Plant disease classification is a vital step in controlling plant diseases. Traditionally, this was achieved using direct and indirect methods like RT-PCR or thermography which are expensive, need a trained expert, and don’t provide a real-time diagnosis. Diseases in plants are caused by various pathogens and identifying them through Machine learning is a major focus of this study as not much work has been done in this field of plant health monitoring and early detection of the pathogen can give insights on how to best combat the infestation. It can also be used to quickly identify the pathogen if a new plant disease is found, and for breeding more pathogen-resistant plants without the need for a trained expert. In this study, we have focused on 5 crop and their disease-causing pathogens, namely Apple Scab in apples, Early Blight in potato, Citrus Greening in oranges, Bacterial Spot in peaches, Late Blight in tomatoes, Spider Mites in Tomatoes, Mosaic Virus in Tomatoes and Yellow Leaf Curl Virus in Tomatoes. We have used computer vision techniques to extract features of the infected leaves from the PlantVillage dataset and achieved promising results using various Supervised Machine Learning techniques such Random Forest, SVM, Naive Bayes, and KNN to classify the pathogen type. The maximum accuracies achieved using Random Forest was 96.6% for Bacteria, 100% for Fungi, 100% for Mites, 93.3% for Mold, and 80% for Viruses. Similarly, for SVM, the maximum accuracies achieved were 96.6% for Bacteria, 93.3% for Fungi, 96.6% for Mites, 100% for Mold, and 93.3% for Viruses. The maximum accuracies achieved using Naive Bayes was 20% for Bacteria, 63.3% for Fungi, 43.3% for Mites, 93.3% for Mold, and 3% for Viruses, and lastly, The maximum accuracies achieved using KNN was 80% for Bacteria, 36.6% for Fungi, 46.6% for Mites, 46.6% for Mold, and 16.6% for Viruses.

1.
FAO, IFAD, UNICEF, WFP and WHO
”, https://www.fao.org/documents/card/en/c/cb4474en, accessed on 2022-01-30
2.
Martinelli
,
F.
,
Scalenghe
,
R.
,
Davino
,
S.
,
Panno
,
S.
,
Scuderi
,
G.
,
Ruisi
,
P.
, … &
Dandekar
,
A. M.
(
2015
).
Advanced methods of plant disease detection. A review
.
Agronomy for Sustainable Development
,
35
(
1
),
1
25
.
3.
Nikita Bisht and Puneet Singh Chauhan
”, https://www.intechopen.com/chapters/74460, accessed on 2022-03-13
4.
Fang
Y
,
Ramasamy
R.P.
Current and Prospective Methods for Plant Disease Detection
.
Biosensors.
2015
;
5
(
3
):
537
561
.
5.
Bock
,
C. H.
,
Barbedo
,
J. G.
,
Del Ponte
,
E. M.
,
Bohnenkamp
,
D.
, &
Mahlein
,
A. K.
(
2020
).
From visual estimates to fully automated sensor-based measurements of plant disease severity: status and challenges for improving accuracy
.
Phytopathology Research
,
2
(
1
),
1
30
.
6.
Verma
,
V. K.
, &
Jain
,
T.
(
2019
). Soft-Computing-Based Approaches for Plant Leaf Disease Detection: Machine-Learning-Based Study. In
Applications of Image Processing and Soft Computing Systems in Agriculture (pp
.
100
113
).
IGI Global
.
7.
Barbedo
,
J. G.A.
(
2016
).
A novel algorithm for semi-automatic segmentation of plant leaf disease symptoms using digital image processing
.
Tropical Plant Pathology
,
41
(
4
),
210
224
.
8.
Sinha
,
A.
, &
Shekhawat
,
R. S.
(
2020
).
Review of image processing approaches for detecting plant diseases
.
IET Image Processing
,
14
(
8
), 1427-1439.
9.
Hughes
,
D.
, &
Salathé
,
M.
(
2015
).
An open access repository of images on plant health to enable the development of mobile disease diagnostics
. arXiv preprint arXiv:1511.08060.
10.
Gardner
,
C. L.
,
da Silva
,
D. R.
,
Pagliai
,
F. A.
,
Pan
,
L.
,
Padgett-Pagliai
,
K. A.
,
Blaustein
,
R. A.
, … &
Gonzalez
,
C. F.
(
2020
).
Assessment of unconventional antimicrobial compounds for the control of ‘Candidatus Liberibacter asiaticus’, the causative agent of citrus greening disease
.
Scientific reports
,
10
(
1
),
1
15
.
12.
Palacio-Bielsa
,
A.
,
Cubero
,
J.
,
Cambra
,
M. A.
,
Collados
,
R.
,
Berruete
,
I. M.
, &
López
,
M. M.
(
2011
).
Development of an efficient real-time quantitative PCR protocol for detection of Xanthomonas arboricola pv. pruni in Prunus species
.
Applied and Environmental Microbiology
,
77
(
1
),
89
97
.
13.
Garita-Cambronero
,
J.
,
Palacio-Bielsa
,
A.
, &
Cubero
,
J.
(
2018
).
Xanthomonas arboricola pv. pruni, causal agent of bacterial spot of stone fruits and almond: its genomic and phenotypic characteristics in the X. arboricola species context
.
Molecular plant pathology
,
19
(
9
),
2053
2065
.
14.
Le Cam
,
B.
,
Sargent
,
D.
,
Gouzy
,
J.
,
Amselem
,
J.
,
Bellanger
,
M. N.
,
Bouchez
,
O.
, … &
Lemaire
,
C.
(
2019
).
Population genome sequencing of the scab fungal species Venturia inaequalis, Venturia pirina, Venturia aucupariae and Venturia asperata
.
G3: Genes, Genomes, Genetics
,
9
(
8
),
2405
2414
.
15.
Agut
,
B.
,
Pastor
,
V.
,
Jaques
,
J. A.
, &
Flors
,
V.
(
2018
).
Can plant defence mechanisms provide new approaches for the sustainable control of the two-spotted spider mite Tetranychus urticae?
.
International journal of molecular sciences
,
19
(
2
),
614
.
16.
van’t Klooster
,
J. W.
,
van den Berg-Velthuis
,
G.
,
van West
,
P.
, &
Govers
,
F.
(
2000
).
tef1, a Phytophthora infestans gene encoding translation elongation factor 1
.
Gene
,
249
(
1-2
),
145
151
.
17.
Broadbent
,
L.
(
1976
).
Epidemiology and control of tomato mosaic virus
.
Annual review of Phytopathology
,
14
(
1
),
75
96
.
18.
Brunt
,
A. A.
(
1986
). Tomato mosaic virus. In
The plant viruses (pp
.
181
204
).
Springer
,
Boston, MA
.
19.
Picó
,
B.
,
Díez
,
M. J.
, &
Nuez
,
F.
(
1996
).
Viral diseases causing the greatest economic losses to the tomato crop. II. The Tomato yellow leaf curl virus—A review
.
Scientia Horticulturae
,
67
(
3-4
),
151
196
.
This content is only available via PDF.
You do not currently have access to this content.