Potato Virus Y Detection in Seed Potatoes Using Deep Learning on Hyperspectral Images

G. Polder*, P.M. Blok, H.A.C. de Villiers, J.M. van der Wolf, J.A.L.M. Kamp

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

3 Citations (Scopus)

Abstract

Virus diseases are of high concern in the cultivation of seed potatoes. Once found inthe field, virus diseased plants lead to declassification or even rejection of the seed lotsresulting in a financial loss. Farmers put in a lot of effort to detect diseased plants andremove virus-diseased plants from the field. Nevertheless, dependent on the cultivar,virus diseased plants can be missed during visual observations in particular in an earlystage of cultivation. Therefore, there is a need for fast and objective disease detection.Early detection of diseased plants with modern vision techniques can significantlyreduce costs. Laboratory experiments in previous years showed that hyperspectral imaging clearly could distinguish healthy from virus infected potato plants. This paper reports on our first real field experiment. A new imaging setup was designed, consisting of a hyperspectral line-scan camera. Hyperspectral images were taken in the field with a line interval of 5 mm. A fully convolutional neural network was adapted for hyperspectral images and trained on two experimental rows in the field. The trained network was validated on two other rows, with different potato cultivars. For three of the four row/date combinations the precision and recall compared to conventional disease assessment exceeded 0.78 and 0.88, respectively. This proves the suitability of this method for real world disease detection.
Original languageEnglish
Article number209
JournalFrontiers in Plant Science
Volume10
DOIs
Publication statusPublished - 7 Mar 2019

Fingerprint

hyperspectral imagery
Potato virus Y
seed tubers
plant viruses
learning
disease detection
image analysis
potatoes
viruses
cultivars
cameras
neural networks
farmers
seeds
methodology

Keywords

  • crop resistance
  • Phenotyping
  • hyperspectral imaging
  • classification
  • Convolutional neural network
  • Solanum tuberosum

Cite this

@article{b9b1481e4639400081c8ee00bc775074,
title = "Potato Virus Y Detection in Seed Potatoes Using Deep Learning on Hyperspectral Images",
abstract = "Virus diseases are of high concern in the cultivation of seed potatoes. Once found inthe field, virus diseased plants lead to declassification or even rejection of the seed lotsresulting in a financial loss. Farmers put in a lot of effort to detect diseased plants andremove virus-diseased plants from the field. Nevertheless, dependent on the cultivar,virus diseased plants can be missed during visual observations in particular in an earlystage of cultivation. Therefore, there is a need for fast and objective disease detection.Early detection of diseased plants with modern vision techniques can significantlyreduce costs. Laboratory experiments in previous years showed that hyperspectral imaging clearly could distinguish healthy from virus infected potato plants. This paper reports on our first real field experiment. A new imaging setup was designed, consisting of a hyperspectral line-scan camera. Hyperspectral images were taken in the field with a line interval of 5 mm. A fully convolutional neural network was adapted for hyperspectral images and trained on two experimental rows in the field. The trained network was validated on two other rows, with different potato cultivars. For three of the four row/date combinations the precision and recall compared to conventional disease assessment exceeded 0.78 and 0.88, respectively. This proves the suitability of this method for real world disease detection.",
keywords = "crop resistance, Phenotyping, hyperspectral imaging, classification, Convolutional neural network, Solanum tuberosum",
author = "G. Polder and P.M. Blok and {de Villiers}, H.A.C. and {van der Wolf}, J.M. and J.A.L.M. Kamp",
year = "2019",
month = "3",
day = "7",
doi = "10.3389/fpls.2019.00209",
language = "English",
volume = "10",
journal = "Frontiers in Plant Science",
issn = "1664-462X",
publisher = "Frontiers",

}

TY - JOUR

T1 - Potato Virus Y Detection in Seed Potatoes Using Deep Learning on Hyperspectral Images

AU - Polder, G.

AU - Blok, P.M.

AU - de Villiers, H.A.C.

AU - van der Wolf, J.M.

AU - Kamp, J.A.L.M.

PY - 2019/3/7

Y1 - 2019/3/7

N2 - Virus diseases are of high concern in the cultivation of seed potatoes. Once found inthe field, virus diseased plants lead to declassification or even rejection of the seed lotsresulting in a financial loss. Farmers put in a lot of effort to detect diseased plants andremove virus-diseased plants from the field. Nevertheless, dependent on the cultivar,virus diseased plants can be missed during visual observations in particular in an earlystage of cultivation. Therefore, there is a need for fast and objective disease detection.Early detection of diseased plants with modern vision techniques can significantlyreduce costs. Laboratory experiments in previous years showed that hyperspectral imaging clearly could distinguish healthy from virus infected potato plants. This paper reports on our first real field experiment. A new imaging setup was designed, consisting of a hyperspectral line-scan camera. Hyperspectral images were taken in the field with a line interval of 5 mm. A fully convolutional neural network was adapted for hyperspectral images and trained on two experimental rows in the field. The trained network was validated on two other rows, with different potato cultivars. For three of the four row/date combinations the precision and recall compared to conventional disease assessment exceeded 0.78 and 0.88, respectively. This proves the suitability of this method for real world disease detection.

AB - Virus diseases are of high concern in the cultivation of seed potatoes. Once found inthe field, virus diseased plants lead to declassification or even rejection of the seed lotsresulting in a financial loss. Farmers put in a lot of effort to detect diseased plants andremove virus-diseased plants from the field. Nevertheless, dependent on the cultivar,virus diseased plants can be missed during visual observations in particular in an earlystage of cultivation. Therefore, there is a need for fast and objective disease detection.Early detection of diseased plants with modern vision techniques can significantlyreduce costs. Laboratory experiments in previous years showed that hyperspectral imaging clearly could distinguish healthy from virus infected potato plants. This paper reports on our first real field experiment. A new imaging setup was designed, consisting of a hyperspectral line-scan camera. Hyperspectral images were taken in the field with a line interval of 5 mm. A fully convolutional neural network was adapted for hyperspectral images and trained on two experimental rows in the field. The trained network was validated on two other rows, with different potato cultivars. For three of the four row/date combinations the precision and recall compared to conventional disease assessment exceeded 0.78 and 0.88, respectively. This proves the suitability of this method for real world disease detection.

KW - crop resistance

KW - Phenotyping

KW - hyperspectral imaging

KW - classification

KW - Convolutional neural network

KW - Solanum tuberosum

U2 - 10.3389/fpls.2019.00209

DO - 10.3389/fpls.2019.00209

M3 - Article

VL - 10

JO - Frontiers in Plant Science

JF - Frontiers in Plant Science

SN - 1664-462X

M1 - 209

ER -