No Cover Image

Journal article 218 views 82 downloads

Temporal Optimisation of Satellite Image‐Based Crop Mapping: A Comparison of Deep Time Series and Semi‐Supervised Time Warping Strategies

Rosie Finnegan, Joseph Metcalfe, Sara Sharifzadeh Orcid Logo, Fabio Caraffini Orcid Logo, Xianghua Xie Orcid Logo, Alberto Hornero, Nicholas W. Synes

IET Computer Vision, Volume: 19, Issue: 1, Start page: e70014

Swansea University Authors: Rosie Finnegan, Joseph Metcalfe, Sara Sharifzadeh Orcid Logo, Fabio Caraffini Orcid Logo, Xianghua Xie Orcid Logo

  • 69047.VOR.pdf

    PDF | Version of Record

    © 2025 The Author(s). IET Computer Vision published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology. This is an open access article under the terms of the Creative Commons Attribution License (CC BY).

    Download (5.04MB)

Check full text

DOI (Published version): 10.1049/cvi2.70014

Abstract

This study presents a novel approach to crop mapping using remotely sensed satellite images. It addresses the significant classification modelling challenges, including (1) the requirements for extensive labelled data and (2) the complex optimisation problem for selection of appropriate temporal win...

Full description

Published in: IET Computer Vision
ISSN: 1751-9632 1751-9640
Published: Wiley 2025
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa69047
Abstract: This study presents a novel approach to crop mapping using remotely sensed satellite images. It addresses the significant classification modelling challenges, including (1) the requirements for extensive labelled data and (2) the complex optimisation problem for selection of appropriate temporal windows in the absence of prior knowledge of cultivation calendars. We compare the lightweight Dynamic Time Warping (DTW) classification method with the heavily supervised Convolutional Neural Network - Long Short-Term Memory (CNN-LSTM) using high-resolution multispectral optical satellite imagery (3 m/pixel). Our approach integrates effective practical preprocessing steps, including data augmentation and a data-driven optimisation strategy for the temporal window, even in the presence of numerous crop classes. Our findings demonstrate that DTW, despite its lower data demands, can match the performance of CNN-LSTM through our effective preprocessing steps while significantly improving runtime. These results demonstrate that both CNN-LSTM and DTW can achieve deployment-level accuracy and underscore the potential of DTW as a viable alternative to more resource-intensive models. The results also prove the effectiveness of temporal windowing for improving runtime and accuracy of a crop classification study, even with no prior knowledge of planting timeframes.
Keywords: computer vision, convolutional neural nets, learning (artificial intelligence), remote sensing, time series
College: Faculty of Science and Engineering
Funders: Funding for this paper was kindly provided by Swansea University EPSRC DTP funding Project Reference EP/W524694/1 and Swansea University Research Excellence funding Scheme of the School of Mathematics and Computer Science.
Issue: 1
Start Page: e70014

Similar Items