Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis

Mariana Belgiu (Corresponding Author), Ovidiu Csillik

Research output: Contribution to journalArticleAcademicpeer-review

101 Citations (Scopus)
264 Downloads (Pure)

Abstract

Efficient methodologies for mapping croplands are an essential condition for the implementation of sustainable agricultural practices and for monitoring crops periodically. The increasing spatial and temporal resolution of globally available satellite images, such as those provided by Sentinel-2, creates new possibilities for generating accurate datasets on available crop types, in ready-to-use vector data format. Existing solutions dedicated to cropland mapping, based on high resolution remote sensing data, are mainly focused on pixel-based analysis of time series data. This paper evaluates how a time-weighted dynamic time warping (TWDTW) method that uses Sentinel-2 time series performs when applied to pixel-based and object-based classifications of various crop types in three different study areas (in Romania, Italy and the USA). The classification outputs were compared to those produced by Random Forest (RF) for both pixel- and object-based image analysis units. The sensitivity of these two methods to the training samples was also evaluated. Object-based TWDTW outperformed pixel-based TWDTW in all three study areas, with overall accuracies ranging between 78.05% and 96.19%; it also proved to be more efficient in terms of computational time. TWDTW achieved comparable classification results to RF in Romania and Italy, but RF achieved better results in the USA, where the classified crops present high intra-class spectral variability. Additionally, TWDTW proved to be less sensitive in relation to the training samples. This is an important asset in areas where inputs for training samples are limited.
Original languageEnglish
Pages (from-to)509-523
Number of pages15
JournalRemote sensing of environment
Volume204
DOIs
Publication statusPublished - Jan 2018

Fingerprint

pixel
Pixels
Crops
Romania
crops
taxonomy
Time series
time series analysis
vector data
Italy
crop
assets
sampling
Image analysis
remote sensing
Remote sensing
methodology
cropland
analysis
image analysis

Keywords

  • ITC-ISI-JOURNAL-ARTICLE
  • ITC-HYBRID

Cite this

@article{ce90971bc037447fa18584d81a41d4d2,
title = "Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis",
abstract = "Efficient methodologies for mapping croplands are an essential condition for the implementation of sustainable agricultural practices and for monitoring crops periodically. The increasing spatial and temporal resolution of globally available satellite images, such as those provided by Sentinel-2, creates new possibilities for generating accurate datasets on available crop types, in ready-to-use vector data format. Existing solutions dedicated to cropland mapping, based on high resolution remote sensing data, are mainly focused on pixel-based analysis of time series data. This paper evaluates how a time-weighted dynamic time warping (TWDTW) method that uses Sentinel-2 time series performs when applied to pixel-based and object-based classifications of various crop types in three different study areas (in Romania, Italy and the USA). The classification outputs were compared to those produced by Random Forest (RF) for both pixel- and object-based image analysis units. The sensitivity of these two methods to the training samples was also evaluated. Object-based TWDTW outperformed pixel-based TWDTW in all three study areas, with overall accuracies ranging between 78.05{\%} and 96.19{\%}; it also proved to be more efficient in terms of computational time. TWDTW achieved comparable classification results to RF in Romania and Italy, but RF achieved better results in the USA, where the classified crops present high intra-class spectral variability. Additionally, TWDTW proved to be less sensitive in relation to the training samples. This is an important asset in areas where inputs for training samples are limited.",
keywords = "ITC-ISI-JOURNAL-ARTICLE, ITC-HYBRID",
author = "Mariana Belgiu and Ovidiu Csillik",
year = "2018",
month = "1",
doi = "10.1016/j.rse.2017.10.005",
language = "English",
volume = "204",
pages = "509--523",
journal = "Remote sensing of environment",
issn = "0034-4257",
publisher = "Elsevier",

}

Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis. / Belgiu, Mariana (Corresponding Author); Csillik, Ovidiu.

In: Remote sensing of environment, Vol. 204, 01.2018, p. 509-523.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis

AU - Belgiu, Mariana

AU - Csillik, Ovidiu

PY - 2018/1

Y1 - 2018/1

N2 - Efficient methodologies for mapping croplands are an essential condition for the implementation of sustainable agricultural practices and for monitoring crops periodically. The increasing spatial and temporal resolution of globally available satellite images, such as those provided by Sentinel-2, creates new possibilities for generating accurate datasets on available crop types, in ready-to-use vector data format. Existing solutions dedicated to cropland mapping, based on high resolution remote sensing data, are mainly focused on pixel-based analysis of time series data. This paper evaluates how a time-weighted dynamic time warping (TWDTW) method that uses Sentinel-2 time series performs when applied to pixel-based and object-based classifications of various crop types in three different study areas (in Romania, Italy and the USA). The classification outputs were compared to those produced by Random Forest (RF) for both pixel- and object-based image analysis units. The sensitivity of these two methods to the training samples was also evaluated. Object-based TWDTW outperformed pixel-based TWDTW in all three study areas, with overall accuracies ranging between 78.05% and 96.19%; it also proved to be more efficient in terms of computational time. TWDTW achieved comparable classification results to RF in Romania and Italy, but RF achieved better results in the USA, where the classified crops present high intra-class spectral variability. Additionally, TWDTW proved to be less sensitive in relation to the training samples. This is an important asset in areas where inputs for training samples are limited.

AB - Efficient methodologies for mapping croplands are an essential condition for the implementation of sustainable agricultural practices and for monitoring crops periodically. The increasing spatial and temporal resolution of globally available satellite images, such as those provided by Sentinel-2, creates new possibilities for generating accurate datasets on available crop types, in ready-to-use vector data format. Existing solutions dedicated to cropland mapping, based on high resolution remote sensing data, are mainly focused on pixel-based analysis of time series data. This paper evaluates how a time-weighted dynamic time warping (TWDTW) method that uses Sentinel-2 time series performs when applied to pixel-based and object-based classifications of various crop types in three different study areas (in Romania, Italy and the USA). The classification outputs were compared to those produced by Random Forest (RF) for both pixel- and object-based image analysis units. The sensitivity of these two methods to the training samples was also evaluated. Object-based TWDTW outperformed pixel-based TWDTW in all three study areas, with overall accuracies ranging between 78.05% and 96.19%; it also proved to be more efficient in terms of computational time. TWDTW achieved comparable classification results to RF in Romania and Italy, but RF achieved better results in the USA, where the classified crops present high intra-class spectral variability. Additionally, TWDTW proved to be less sensitive in relation to the training samples. This is an important asset in areas where inputs for training samples are limited.

KW - ITC-ISI-JOURNAL-ARTICLE

KW - ITC-HYBRID

UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2018/isi/belgiu_sen.pdf

U2 - 10.1016/j.rse.2017.10.005

DO - 10.1016/j.rse.2017.10.005

M3 - Article

VL - 204

SP - 509

EP - 523

JO - Remote sensing of environment

JF - Remote sensing of environment

SN - 0034-4257

ER -