Paddy rice and winter wheat identification based on Sentinel-2 imagery with semi-supervised learning (SSL)

ABSTRACT Accurately identifying the distribution of various crops and mapping their spatial patterns is crucial for modern agricultural monitoring and management. Deep learning methods have shown strong performance in crop mapping using remote sensing, but their effectiveness is often limited by the amount of samples. This study leverages Sentinel-2 satellite remote sensing imagery and two deep learning-based Semi-Supervised Learning (SSL) strategies (ST and ST++) to accurately identify two major crops – paddy rice and winter wheat – in the Bengbu region of Anhui Province, China. The study also compares the improvements offered by fully supervised and semi-supervised methods for crop identification. Results demonstrate that increasing the amount of labelled data enhances the performance of fully supervised models. Additionally, both SSL strategies further improve the identification results of fully supervised models, albeit with an increased training time. Notably, the ST strategy is particularly effective when using 1–10% of labelled data, capturing finer details that fully supervised learning may overlook. The ST++ strategy further refines model performance metrics compared to ST. Specifically, when using 5%, 30% and 40% labelled data, the crop identification results from SSL strategies for paddy rice and winter wheat are much closer to ground truth samples. This study highlights the potential of SSL strategies in enhancing crop identification tasks using limited labelled data.

• Publication year

  2025

• Venue

  International Journal of Remote Sensing

• Publication date

  2025-11-10

• Fields of study

  Not labeled

• Identifiers
  DOI 10.1080/01431161.2025.2582877
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Efficient weed segmentation in maize fields: A semi-supervised approach for precision weed management with reduced annotation overhead

  2025cited by this paper
• Remote sensing and artificial intelligence: revolutionizing pest management in agriculture

  2025cited by this paper
• A Weakly Supervised and Self-Supervised Learning Approach for Semantic Segmentation of Land Cover in Satellite Images with National Forest Inventory Data

  2025cited by this paper
• A difference enhancement and class-aware rebalancing semi-supervised network for cropland semantic change detection

  2025cited by this paper
• Segmenting vegetation from UAV images via spectral reconstruction in complex field environments

  2025cited by this paper
• A Deep Learning Based Platform for Remote Sensing Images Change Detection Integrating Crowdsourcing and Active Learning

  2024cited by this paper
• Editorial: Precise chemical application technology for horticultural crops

  2024cited by this paper
• Advancing perturbation space expansion based on information fusion for semi-supervised remote sensing image semantic segmentation

  2024cited by this paper
• Semantic Segmentation of Remotely Sensed Images using Multisource Data: An Experimental Analysis

  2024cited by this paper
• Pixel to practice: multi-scale image data for calibrating remote-sensing-based winter wheat monitoring methods

  2024cited by this paper
• Semi-Supervised Semantic Change Detection Based on Weak-to-Strong Consistency

  2024cited by this paper
• Bi-Objective Crop Mapping from Sentinel-2 Images Based on Multiple Deep Learning Networks

  2023cited by this paper
• Crop Classification in Mountainous Areas Using Object-Oriented Methods and Multi-Source Data: A Case Study of Xishui County, China

  2023cited by this paper
• Image patch-based deep learning approach for crop and weed recognition

  2023cited by this paper
• Classification of Crops through Self-Supervised Decomposition for Transfer Learning

  2023cited by this paper
• Identifying Winter Wheat Using Landsat Data Based on Deep Learning Algorithms in the North China Plain

  2023cited by this paper
• Application of Deep Learning in Multitemporal Remote Sensing Image Classification

  2023cited by this paper
• Cost-efficient information extraction from massive remote sensing data: When weakly supervised deep learning meets remote sensing big data

  2023cited by this paper
• Global crop calendars of maize and wheat in the framework of the WorldCereal project

  2022cited by this paper
• Multi-Season Phenology Mapping of Nile Delta Croplands Using Time Series of Sentinel-2 and Landsat 8 Green LAI

  2022cited by this paper
• Cropping Patterns of Annual Crops: A Remote Sensing Review

  2022cited by this paper
• Multi-Crop Classification Using Feature Selection-Coupled Machine Learning Classifiers Based on Spectral, Textural and Environmental Features

  2022cited by this paper
• Weakly supervised high spatial resolution land cover mapping based on self-training with weighted pseudo-labels

  2022cited by this paper
• Unsupervised Domain Adaptation with Adversarial Self-Training for Crop Classification Using Remote Sensing Images

  2022cited by this paper
• Self-Supervised Encoders Are Better Transfer Learners in Remote Sensing Applications

  2022cited by this paper
• MSResNet: Multiscale Residual Network via Self-Supervised Learning for Water-Body Detection in Remote Sensing Imagery

  2021cited by this paper
• Early- and in-season crop type mapping without current-year ground truth: generating labels from historical information via a topology-based approach

  2021cited by this paper
• Monitoring the summer flooding in the Poyang Lake area of China in 2020 based on Sentinel-1 data and multiple convolutional neural networks

  2021cited by this paper
• Mapping cropping intensity in Huaihe basin using phenology algorithm, all Sentinel-2 and Landsat images in Google Earth Engine

  2021cited by this paper
• Weed Detection in Agricultural fields using Deep Learning Process

  2021cited by this paper
• Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision

  2021cited by this paper
• A New Method for Winter Wheat Mapping Based on Spectral Reconstruction Technology

  2021cited by this paper
• The 10-m crop type maps in Northeast China during 2017–2019

  2021cited by this paper
• ChinaCropPhen1km: a high-resolution crop phenological dataset for three staple crops in China during 2000–2015 based on leaf area index (LAI) products

  2020cited by this paper
• FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

  2020cited by this paper
• Mapping Winter Wheat in North China Using Sentinel 2A/B Data: A Method Based on Phenology-Time Weighted Dynamic Time Warping

  2020cited by this paper
• Self-Training With Noisy Student Improves ImageNet Classification

  2019cited by this paper
• Crop classification in a heterogeneous agricultural environment using ensemble classifiers and single-date Sentinel-2A imagery

  2019cited by this paper
• Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learning

  2019cited by this paper
• A survey on Image Data Augmentation for Deep Learning

  2019cited by this paper
• Focal Loss for Dense Object Detection

  2017cited by this paper
• Pyramid Scene Parsing Network

  2016cited by this paper
• Efficiency Assessment of Multitemporal C-Band Radarsat-2 Intensity and Landsat-8 Surface Reflectance Satellite Imagery for Crop Classification in Ukraine

  2016cited by this paper
• Detection and Modeling of Vegetation Phenology Spatiotemporal Characteristics in the Middle Part of the Huai River Region in China

  2015cited by this paper
• Responses of a rice–wheat rotation agroecosystem to experimental warming

  2013cited by this paper
• Climate Change and its Impact on Water Resources in the Huai River Basin

  2012cited by this paper

• No citing papers are available for this paper.