Mapping of Weed Management Methods in Orchards using Sentinel-2 and PlanetScope Data

2025-04-28

AI Summary

Traditional ground surveys for orchard weed management—such as mowing, tillage, herbicide application, and no intervention—are costly, temporally delayed, and spatially limited. To address these limitations, this study integrates time-series optical imagery from Sentinel-2 and PlanetScope, proposing a phenology-aware temporal feature modeling method tailored to orchard structure and agricultural cycles. We further develop an ensemble learning framework combining Random Forest and Convolutional Neural Networks for fine-grained classification of management practices. This work presents the first deep fusion of dual-source optical time-series data for orchard weed management classification and innovatively introduces dynamic vegetation index modeling to enhance phenological discriminability. Evaluated on real-world orchard sites, the method achieves a mean classification accuracy of 89.3%. It significantly improves monitoring efficiency, spatiotemporal coverage, and scalability, thereby providing robust remote sensing support for agricultural policy compliance assessment, ecological benefit quantification, and sustainable management decision-making.

Technology Category

• Machine Learning: Ensemble Methods
• Computer Vision: Remote Sensing / Geospatial AI
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Effective weed management is crucial for improving agricultural productivity, as weeds compete with crops for vital resources like nutrients and water. Accurate maps of weed management methods are essential for policymakers to assess farmer practices, evaluate impacts on vegetation health, biodiversity, and climate, as well as ensure compliance with policies and subsidies. However, monitoring weed management methods is challenging as commonly rely on on-ground field surveys, which are often costly, time-consuming and subject to delays. In order to tackle this problem, we leverage Earth Observation (EO) data and Machine Learning (ML). Specifically, we developed an ML approach for mapping four distinct weed management methods (Mowing, Tillage, Chemical-spraying, and No practice) in orchards using satellite image time series (SITS) data from two different sources: Sentinel-2 (S2) and PlanetScope (PS). The findings demonstrate the potential of ML-driven remote sensing to enhance the efficiency and accuracy of weed management mapping in orchards.

Problem

• Mapping weed management methods in orchards using satellite data
• Overcoming costly and slow field surveys with remote sensing
• Classifying four weed management practices via machine learning

Innovation

• Uses Sentinel-2 and PlanetScope satellite data
• Applies Machine Learning to classify weed methods
• Analyzes time series for orchard weed management