Time series of Sentinel-1 and Sentinel-2 imagery for parcel-based crop-type classification using Random Forest algorithm and Google Earth Engine
Ghorbanian, Arsalan ; Zaghian, Soheil ; Ahmadi, Seyed Ali ; Mohammadzadeh, Ali and Jamali, Sadegh LU
(2025)
p.303-319
- Abstract
Precise statistics about crop types and productions can support establishing more efficient decisions in the food security framework. The applicability of remote sensing imagery has been confirmed worldwide in crop-related studies. In this chapter, time series Sentinel-1 and Sentinel-2 images were integrated, along with crop inventory, to generate crop-type maps in Imperial County, CA, USA, within Google Colab and Google Earth Engine (GEE). Crop parcels were considered as the classification unit, and monthly vegetation indices and backscattering coefficients, along with crop type information, were fed to two Random Forest classifiers to generate two maps concerning major crop classes and crop subclasses. The first map was produced in... (More)
Precise statistics about crop types and productions can support establishing more efficient decisions in the food security framework. The applicability of remote sensing imagery has been confirmed worldwide in crop-related studies. In this chapter, time series Sentinel-1 and Sentinel-2 images were integrated, along with crop inventory, to generate crop-type maps in Imperial County, CA, USA, within Google Colab and Google Earth Engine (GEE). Crop parcels were considered as the classification unit, and monthly vegetation indices and backscattering coefficients, along with crop type information, were fed to two Random Forest classifiers to generate two maps concerning major crop classes and crop subclasses. The first map was produced in five major crop classes with an F1-score and Overall Accuracy (OA) of 0.86% and 95.13%, respectively. Similarly, the second map containing 12 crop classes had an F1-score of 0.82 and an Overall Accuracy of 90.72%. The multitemporal and multimodal framework led to achieving high classification accuracies and proved its capability for operational practices.
(Less)
- author
- Ghorbanian, Arsalan
; Zaghian, Soheil
; Ahmadi, Seyed Ali
; Mohammadzadeh, Ali
and Jamali, Sadegh
LU
- organization
- publishing date
- 2025-01-01
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Classification, Crop, Google Earth Engine, Random Forest, Sentinel, Time series, USA
- host publication
- Google Earth Engine and Artificial Intelligence for Earth Observation : Algorithms and Sustainable Applications - Algorithms and Sustainable Applications
- pages
- 17 pages
- publisher
- Elsevier
- external identifiers
-
- scopus:105011231204
- ISBN
- 9780443273735
- 9780443273728
- DOI
- 10.1016/B978-0-443-27372-8.00008-8
- language
- English
- LU publication?
- yes
- additional info
- Publisher Copyright: © 2025 Elsevier Ltd. All rights reserved.
- id
- 490cc222-e686-420b-b099-3e74487e9ec1
- date added to LUP
- 2026-01-12 15:00:56
- date last changed
- 2026-01-12 15:01:52
@inbook{490cc222-e686-420b-b099-3e74487e9ec1,
abstract = {{<p>Precise statistics about crop types and productions can support establishing more efficient decisions in the food security framework. The applicability of remote sensing imagery has been confirmed worldwide in crop-related studies. In this chapter, time series Sentinel-1 and Sentinel-2 images were integrated, along with crop inventory, to generate crop-type maps in Imperial County, CA, USA, within Google Colab and Google Earth Engine (GEE). Crop parcels were considered as the classification unit, and monthly vegetation indices and backscattering coefficients, along with crop type information, were fed to two Random Forest classifiers to generate two maps concerning major crop classes and crop subclasses. The first map was produced in five major crop classes with an F1-score and Overall Accuracy (OA) of 0.86% and 95.13%, respectively. Similarly, the second map containing 12 crop classes had an F1-score of 0.82 and an Overall Accuracy of 90.72%. The multitemporal and multimodal framework led to achieving high classification accuracies and proved its capability for operational practices.</p>}},
author = {{Ghorbanian, Arsalan and Zaghian, Soheil and Ahmadi, Seyed Ali and Mohammadzadeh, Ali and Jamali, Sadegh}},
booktitle = {{Google Earth Engine and Artificial Intelligence for Earth Observation : Algorithms and Sustainable Applications}},
isbn = {{9780443273735}},
keywords = {{Classification; Crop; Google Earth Engine; Random Forest; Sentinel; Time series; USA}},
language = {{eng}},
month = {{01}},
pages = {{303--319}},
publisher = {{Elsevier}},
title = {{Time series of Sentinel-1 and Sentinel-2 imagery for parcel-based crop-type classification using Random Forest algorithm and Google Earth Engine}},
url = {{http://dx.doi.org/10.1016/B978-0-443-27372-8.00008-8}},
doi = {{10.1016/B978-0-443-27372-8.00008-8}},
year = {{2025}},
}