Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation

Generating composite images from satellite data is crucial for crop mapping over defined periods. However, producing reliable composites for cropland segmentation presents challenges, particularly in maintaining temporal coherence and preserving key phenological stages in time series data. This stud...

Full description

Saved in:
Bibliographic Details
Main Authors: Reza Maleki, Falin Wu, Guoxin Qu, Amel Oubara, Gongliu Yang
Format: Article
Language:English
Published: Elsevier 2025-09-01
Series:International Journal of Applied Earth Observations and Geoinformation
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1569843225004510
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849225346680881152
author Reza Maleki
Falin Wu
Guoxin Qu
Amel Oubara
Gongliu Yang
author_facet Reza Maleki
Falin Wu
Guoxin Qu
Amel Oubara
Gongliu Yang
author_sort Reza Maleki
collection DOAJ
description Generating composite images from satellite data is crucial for crop mapping over defined periods. However, producing reliable composites for cropland segmentation presents challenges, particularly in maintaining temporal coherence and preserving key phenological stages in time series data. This study proposes a compositing method that improves temporal coherence for tracking phenological stages in deep learning–based cropland segmentation. The compositing method integrates the near–infrared to blue band reflectance ratio with a Gaussian weighting function to prioritize pixel selection based on temporal proximity to the center of the target month. Sentinel–2 monthly time series composites were generated using Google Earth Engine and evaluated through proximity analysis to assess pixel distribution within the target month and correlations with consecutive months. The performance of deep learning models trained on these composites was further assessed by comparing their segmentation results. To evaluate generalizability, the method was applied across various study areas and across different crop types and environmental conditions. The results consistently show that proposed method outperforms other techniques in preserving temporal continuity, reducing cloud–related noise, and maintaining the coherence necessary for deep learning models to effectively track crop growth patterns.
format Article
id doaj-art-9a3a15b9edb94e75b91decbfc70955ee
institution Kabale University
issn 1569-8432
language English
publishDate 2025-09-01
publisher Elsevier
record_format Article
series International Journal of Applied Earth Observations and Geoinformation
spelling doaj-art-9a3a15b9edb94e75b91decbfc70955ee2025-08-25T04:14:11ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322025-09-0114310480410.1016/j.jag.2025.104804Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentationReza Maleki0Falin Wu1Guoxin Qu2Amel Oubara3Gongliu Yang4SNARS Laboratory, School of Instrumentation and Optoelectronic Engineering, Beihang University, Beijing 100191, ChinaSNARS Laboratory, School of Instrumentation and Optoelectronic Engineering, Beihang University, Beijing 100191, China; Corresponding author.Department of System Design, Beijing System Design Institute of Electro–Mechanic Engineering, Beijing 100854, ChinaSNARS Laboratory, School of Instrumentation and Optoelectronic Engineering, Beihang University, Beijing 100191, ChinaSchool of Mechanical Engineering, Zhejiang University, Hangzhou 310030, ChinaGenerating composite images from satellite data is crucial for crop mapping over defined periods. However, producing reliable composites for cropland segmentation presents challenges, particularly in maintaining temporal coherence and preserving key phenological stages in time series data. This study proposes a compositing method that improves temporal coherence for tracking phenological stages in deep learning–based cropland segmentation. The compositing method integrates the near–infrared to blue band reflectance ratio with a Gaussian weighting function to prioritize pixel selection based on temporal proximity to the center of the target month. Sentinel–2 monthly time series composites were generated using Google Earth Engine and evaluated through proximity analysis to assess pixel distribution within the target month and correlations with consecutive months. The performance of deep learning models trained on these composites was further assessed by comparing their segmentation results. To evaluate generalizability, the method was applied across various study areas and across different crop types and environmental conditions. The results consistently show that proposed method outperforms other techniques in preserving temporal continuity, reducing cloud–related noise, and maintaining the coherence necessary for deep learning models to effectively track crop growth patterns.http://www.sciencedirect.com/science/article/pii/S1569843225004510Satellite image compositingSentinel–2 time seriesGoogle earth engineCropland segmentationDeep learningCrop mapping
spellingShingle Reza Maleki
Falin Wu
Guoxin Qu
Amel Oubara
Gongliu Yang
Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
International Journal of Applied Earth Observations and Geoinformation
Satellite image compositing
Sentinel–2 time series
Google earth engine
Cropland segmentation
Deep learning
Crop mapping
title Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
title_full Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
title_fullStr Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
title_full_unstemmed Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
title_short Enhancing satellite image compositing with temporal proximity weighting for deep learning–based cropland segmentation
title_sort enhancing satellite image compositing with temporal proximity weighting for deep learning based cropland segmentation
topic Satellite image compositing
Sentinel–2 time series
Google earth engine
Cropland segmentation
Deep learning
Crop mapping
url http://www.sciencedirect.com/science/article/pii/S1569843225004510
work_keys_str_mv AT rezamaleki enhancingsatelliteimagecompositingwithtemporalproximityweightingfordeeplearningbasedcroplandsegmentation
AT falinwu enhancingsatelliteimagecompositingwithtemporalproximityweightingfordeeplearningbasedcroplandsegmentation
AT guoxinqu enhancingsatelliteimagecompositingwithtemporalproximityweightingfordeeplearningbasedcroplandsegmentation
AT ameloubara enhancingsatelliteimagecompositingwithtemporalproximityweightingfordeeplearningbasedcroplandsegmentation
AT gongliuyang enhancingsatelliteimagecompositingwithtemporalproximityweightingfordeeplearningbasedcroplandsegmentation