Crowd-Driven and Automated Mapping of Field Boundaries in Highly Fragmented Agricultural Landscapes of Ethiopia with Very High Spatial Resolution Imagery

M.T. Marshall (Corresponding Author), S. Crommelinck, D. Kohli, Christoph Perger, M.Y. Yang, Aniruddha Ghosh, Steffen Fritz, C.A.J.M. de Bie, A.D. Nelson

Research output: Contribution to journalArticleAcademicpeer-review

7 Downloads (Pure)

Abstract

Mapping the extent and location of field boundaries is critical to food security analysis but remains problematic in the Global South where such information is needed the most. The difficulty is due primarily to fragmentation in the landscape, small farm sizes, and irregular farm boundaries. Very high-resolution satellite imagery affords an opportunity to delineate such fields, but the challenge remains of determining such boundaries in a systematic and accurate way. In this paper, we compare a new crowd-driven manual digitization tool (Crop Land Extent) with two semi-automated methods (contour detection and multi-resolution segmentation) to determine farm boundaries from WorldView imagery in highly fragmented agricultural landscapes of Ethiopia. More than 7000 one square-kilometer image tiles were used for the analysis. The three methods were assessed using quantitative completeness and spatial correctness. Contour detection tended to under-segment when compared to manual digitization, resulting in better performance for larger (approaching 1 ha) sized fields. Multi-resolution segmentation on the other hand, tended to over-segment, resulting in better performance for small fields. Neither semi-automated method in their current realizations however are suitable for field boundary mapping in highly fragmented landscapes. Crowd-driven manual digitization is promising, but requires more oversight, quality control, and training than the current workflow could allow. View Full-Text
Original languageEnglish
Article number2082
JournalRemote sensing
Volume11
DOIs
Publication statusPublished - 5 Sep 2019

Fingerprint

digitization
field margin
spatial resolution
imagery
agricultural land
segmentation
farm
farm size
food security
satellite imagery
quality control
fragmentation
crop
method
analysis
detection

Keywords

  • agriculture
  • cropland
  • food security
  • image segmentation
  • object detection
  • crowdsourcing
  • Remote sensing
  • WorldView-2
  • ITC-ISI-JOURNAL-ARTICLE
  • ITC-GOLD

Cite this

@article{27847b72510b4520975586a957a65e12,
title = "Crowd-Driven and Automated Mapping of Field Boundaries in Highly Fragmented Agricultural Landscapes of Ethiopia with Very High Spatial Resolution Imagery",
abstract = "Mapping the extent and location of field boundaries is critical to food security analysis but remains problematic in the Global South where such information is needed the most. The difficulty is due primarily to fragmentation in the landscape, small farm sizes, and irregular farm boundaries. Very high-resolution satellite imagery affords an opportunity to delineate such fields, but the challenge remains of determining such boundaries in a systematic and accurate way. In this paper, we compare a new crowd-driven manual digitization tool (Crop Land Extent) with two semi-automated methods (contour detection and multi-resolution segmentation) to determine farm boundaries from WorldView imagery in highly fragmented agricultural landscapes of Ethiopia. More than 7000 one square-kilometer image tiles were used for the analysis. The three methods were assessed using quantitative completeness and spatial correctness. Contour detection tended to under-segment when compared to manual digitization, resulting in better performance for larger (approaching 1 ha) sized fields. Multi-resolution segmentation on the other hand, tended to over-segment, resulting in better performance for small fields. Neither semi-automated method in their current realizations however are suitable for field boundary mapping in highly fragmented landscapes. Crowd-driven manual digitization is promising, but requires more oversight, quality control, and training than the current workflow could allow. View Full-Text",
keywords = "agriculture, cropland, food security, image segmentation, object detection, crowdsourcing, Remote sensing, WorldView-2, ITC-ISI-JOURNAL-ARTICLE, ITC-GOLD",
author = "M.T. Marshall and S. Crommelinck and D. Kohli and Christoph Perger and M.Y. Yang and Aniruddha Ghosh and Steffen Fritz and {de Bie}, C.A.J.M. and A.D. Nelson",
year = "2019",
month = "9",
day = "5",
doi = "10.3390/rs11182082",
language = "English",
volume = "11",
journal = "Remote sensing",
issn = "2072-4292",
publisher = "Multidisciplinary Digital Publishing Institute",

}

TY - JOUR

T1 - Crowd-Driven and Automated Mapping of Field Boundaries in Highly Fragmented Agricultural Landscapes of Ethiopia with Very High Spatial Resolution Imagery

AU - Marshall, M.T.

AU - Crommelinck, S.

AU - Kohli, D.

AU - Perger, Christoph

AU - Yang, M.Y.

AU - Ghosh, Aniruddha

AU - Fritz, Steffen

AU - de Bie, C.A.J.M.

AU - Nelson, A.D.

PY - 2019/9/5

Y1 - 2019/9/5

N2 - Mapping the extent and location of field boundaries is critical to food security analysis but remains problematic in the Global South where such information is needed the most. The difficulty is due primarily to fragmentation in the landscape, small farm sizes, and irregular farm boundaries. Very high-resolution satellite imagery affords an opportunity to delineate such fields, but the challenge remains of determining such boundaries in a systematic and accurate way. In this paper, we compare a new crowd-driven manual digitization tool (Crop Land Extent) with two semi-automated methods (contour detection and multi-resolution segmentation) to determine farm boundaries from WorldView imagery in highly fragmented agricultural landscapes of Ethiopia. More than 7000 one square-kilometer image tiles were used for the analysis. The three methods were assessed using quantitative completeness and spatial correctness. Contour detection tended to under-segment when compared to manual digitization, resulting in better performance for larger (approaching 1 ha) sized fields. Multi-resolution segmentation on the other hand, tended to over-segment, resulting in better performance for small fields. Neither semi-automated method in their current realizations however are suitable for field boundary mapping in highly fragmented landscapes. Crowd-driven manual digitization is promising, but requires more oversight, quality control, and training than the current workflow could allow. View Full-Text

AB - Mapping the extent and location of field boundaries is critical to food security analysis but remains problematic in the Global South where such information is needed the most. The difficulty is due primarily to fragmentation in the landscape, small farm sizes, and irregular farm boundaries. Very high-resolution satellite imagery affords an opportunity to delineate such fields, but the challenge remains of determining such boundaries in a systematic and accurate way. In this paper, we compare a new crowd-driven manual digitization tool (Crop Land Extent) with two semi-automated methods (contour detection and multi-resolution segmentation) to determine farm boundaries from WorldView imagery in highly fragmented agricultural landscapes of Ethiopia. More than 7000 one square-kilometer image tiles were used for the analysis. The three methods were assessed using quantitative completeness and spatial correctness. Contour detection tended to under-segment when compared to manual digitization, resulting in better performance for larger (approaching 1 ha) sized fields. Multi-resolution segmentation on the other hand, tended to over-segment, resulting in better performance for small fields. Neither semi-automated method in their current realizations however are suitable for field boundary mapping in highly fragmented landscapes. Crowd-driven manual digitization is promising, but requires more oversight, quality control, and training than the current workflow could allow. View Full-Text

KW - agriculture

KW - cropland

KW - food security

KW - image segmentation

KW - object detection

KW - crowdsourcing

KW - Remote sensing

KW - WorldView-2

KW - ITC-ISI-JOURNAL-ARTICLE

KW - ITC-GOLD

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

U2 - 10.3390/rs11182082

DO - 10.3390/rs11182082

M3 - Article

VL - 11

JO - Remote sensing

JF - Remote sensing

SN - 2072-4292

M1 - 2082

ER -