Vegetable classification in Indonesia using Dynamic Time Warping of Sentinel-1A dual polarization SAR time series

Mengmeng Li (Corresponding Author), W. Bijker

Research output: Contribution to journalArticleAcademicpeer-review

4 Downloads (Pure)

Abstract

This study investigates the potential of Sentinel-1A (S1A) dual polarization SAR time series data for vegetable
classification in Indonesia. Vegetables are characterized by the temporal changes of observables extracted from
time series of S1A data. We extracted observables regarding both backscatter (VH and VV) coefficients and
decomposition features (i.e., entropy, angle, and anisotropy). The vegetable classification is based on a timeweighted
Dynamic Time Warping dissimilarity measure that is calculated with SPRING strategy for subsequence
searching, referred to as twDTWS. This study focuses on three main vegetable types widely planted in Indonesia,
namely chili, tomato, and cucumber. We conducted vegetable classification in two areas, Malang and Lampung,
using time series of S1A data covering the dry season in 2017. Our results show that the twDTWS method
provides a promising means to classify vegetables using time series of S1A data for the dry season, while the
features decomposed from dual polarization S1A data have little influence on the classification accuracy.
Moreover, the twDTWS method with query sequences (namely reference temporal profiles) defined on the
Malang dataset produced an overall accuracy of 0.80 for the classification of chili and cucumber from the
Lampung dataset when the query sequences correspond to the growth cycles of vegetables. The variation in the
length (i.e., the number of observations) of query sequences can affect the classification accuracy. We conclude
that the twDTWS method has a high potential for classifying vegetables in different areas when constructing the
query sequences of vegetables based on their growth cycles.
Original languageEnglish
Pages (from-to)268-280
Number of pages13
JournalInternational Journal of Applied Earth Observation and Geoinformation (JAG)
Volume78
Issue numberJune
Early online date28 Feb 2019
DOIs
Publication statusPublished - 1 Jun 2019

Fingerprint

Vegetables
vegetable
Time series
synthetic aperture radar
polarization
time series
Polarization
dry season
backscatter
entropy
Anisotropy
anisotropy
Entropy

Keywords

  • Vegetable classification
  • Time series classification
  • Sentinel-1A
  • Dual polarization SAR
  • Dynamic Time Warping
  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{b933123fa22c4018af73c63559cf6cc7,
title = "Vegetable classification in Indonesia using Dynamic Time Warping of Sentinel-1A dual polarization SAR time series",
abstract = "This study investigates the potential of Sentinel-1A (S1A) dual polarization SAR time series data for vegetableclassification in Indonesia. Vegetables are characterized by the temporal changes of observables extracted fromtime series of S1A data. We extracted observables regarding both backscatter (VH and VV) coefficients anddecomposition features (i.e., entropy, angle, and anisotropy). The vegetable classification is based on a timeweightedDynamic Time Warping dissimilarity measure that is calculated with SPRING strategy for subsequencesearching, referred to as twDTWS. This study focuses on three main vegetable types widely planted in Indonesia,namely chili, tomato, and cucumber. We conducted vegetable classification in two areas, Malang and Lampung,using time series of S1A data covering the dry season in 2017. Our results show that the twDTWS methodprovides a promising means to classify vegetables using time series of S1A data for the dry season, while thefeatures decomposed from dual polarization S1A data have little influence on the classification accuracy.Moreover, the twDTWS method with query sequences (namely reference temporal profiles) defined on theMalang dataset produced an overall accuracy of 0.80 for the classification of chili and cucumber from theLampung dataset when the query sequences correspond to the growth cycles of vegetables. The variation in thelength (i.e., the number of observations) of query sequences can affect the classification accuracy. We concludethat the twDTWS method has a high potential for classifying vegetables in different areas when constructing thequery sequences of vegetables based on their growth cycles.",
keywords = "Vegetable classification, Time series classification, Sentinel-1A, Dual polarization SAR, Dynamic Time Warping, ITC-ISI-JOURNAL-ARTICLE",
author = "Mengmeng Li and W. Bijker",
year = "2019",
month = "6",
day = "1",
doi = "10.1016/j.jag.2019.01.009",
language = "English",
volume = "78",
pages = "268--280",
journal = "International Journal of Applied Earth Observation and Geoinformation (JAG)",
issn = "1569-8432",
publisher = "Elsevier",
number = "June",

}

Vegetable classification in Indonesia using Dynamic Time Warping of Sentinel-1A dual polarization SAR time series. / Li, Mengmeng (Corresponding Author); Bijker, W.

In: International Journal of Applied Earth Observation and Geoinformation (JAG), Vol. 78, No. June, 01.06.2019, p. 268-280.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Vegetable classification in Indonesia using Dynamic Time Warping of Sentinel-1A dual polarization SAR time series

AU - Li, Mengmeng

AU - Bijker, W.

PY - 2019/6/1

Y1 - 2019/6/1

N2 - This study investigates the potential of Sentinel-1A (S1A) dual polarization SAR time series data for vegetableclassification in Indonesia. Vegetables are characterized by the temporal changes of observables extracted fromtime series of S1A data. We extracted observables regarding both backscatter (VH and VV) coefficients anddecomposition features (i.e., entropy, angle, and anisotropy). The vegetable classification is based on a timeweightedDynamic Time Warping dissimilarity measure that is calculated with SPRING strategy for subsequencesearching, referred to as twDTWS. This study focuses on three main vegetable types widely planted in Indonesia,namely chili, tomato, and cucumber. We conducted vegetable classification in two areas, Malang and Lampung,using time series of S1A data covering the dry season in 2017. Our results show that the twDTWS methodprovides a promising means to classify vegetables using time series of S1A data for the dry season, while thefeatures decomposed from dual polarization S1A data have little influence on the classification accuracy.Moreover, the twDTWS method with query sequences (namely reference temporal profiles) defined on theMalang dataset produced an overall accuracy of 0.80 for the classification of chili and cucumber from theLampung dataset when the query sequences correspond to the growth cycles of vegetables. The variation in thelength (i.e., the number of observations) of query sequences can affect the classification accuracy. We concludethat the twDTWS method has a high potential for classifying vegetables in different areas when constructing thequery sequences of vegetables based on their growth cycles.

AB - This study investigates the potential of Sentinel-1A (S1A) dual polarization SAR time series data for vegetableclassification in Indonesia. Vegetables are characterized by the temporal changes of observables extracted fromtime series of S1A data. We extracted observables regarding both backscatter (VH and VV) coefficients anddecomposition features (i.e., entropy, angle, and anisotropy). The vegetable classification is based on a timeweightedDynamic Time Warping dissimilarity measure that is calculated with SPRING strategy for subsequencesearching, referred to as twDTWS. This study focuses on three main vegetable types widely planted in Indonesia,namely chili, tomato, and cucumber. We conducted vegetable classification in two areas, Malang and Lampung,using time series of S1A data covering the dry season in 2017. Our results show that the twDTWS methodprovides a promising means to classify vegetables using time series of S1A data for the dry season, while thefeatures decomposed from dual polarization S1A data have little influence on the classification accuracy.Moreover, the twDTWS method with query sequences (namely reference temporal profiles) defined on theMalang dataset produced an overall accuracy of 0.80 for the classification of chili and cucumber from theLampung dataset when the query sequences correspond to the growth cycles of vegetables. The variation in thelength (i.e., the number of observations) of query sequences can affect the classification accuracy. We concludethat the twDTWS method has a high potential for classifying vegetables in different areas when constructing thequery sequences of vegetables based on their growth cycles.

KW - Vegetable classification

KW - Time series classification

KW - Sentinel-1A

KW - Dual polarization SAR

KW - Dynamic Time Warping

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2019/isi/li_veg.pdf

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.jag.2019.01.009

U2 - 10.1016/j.jag.2019.01.009

DO - 10.1016/j.jag.2019.01.009

M3 - Article

VL - 78

SP - 268

EP - 280

JO - International Journal of Applied Earth Observation and Geoinformation (JAG)

JF - International Journal of Applied Earth Observation and Geoinformation (JAG)

SN - 1569-8432

IS - June

ER -