Monitoring tar spot disease in corn at different canopy and temporal levels using aerial multispectral imaging and machine learning

Aug 5, 2026

Upload by Jiven

Aerial multispectral imaging combined with machine learning supports monitoring tar spot disease in corn across different canopy heights and time periods. The study uses unmanned aircraft systems to capture multispectral images and visual disease severity measurements in micro-plots across multiple time points over two years. Image-derived features from ortho-mosaics, including reflectance bands and vegetation indices, train disease quantification models. The resulting models estimate severity with strong agreement across canopy levels and enable digital phenotyping when severity is low, supporting evaluation of management tactics.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Introduction

Tar spot background and disease impact

Methods

UAS multispectral imaging and visual assessment

Feature extraction and machine learning models

Results and Discussion

Disease severity estimation across canopy and years

Modeling epidemiological parameters

Implications for early monitoring and management evaluation

TYPE Original Research PUBLISHED 23 January 2023 DOI 10.3389/fpls.2022.1077403 OPEN ACCESS EDITED BY Yanan Wang, Agricultural University of Hebei, China REVIEWED BY Darko Jevremovic´, Fruit Research Institute, Serbia Michele Pisante, University of Teramo, Italy *CORRESPONDENCE C. D. Cruz [cruz113@purdue.edu](cruz113@purdue.edu) SPECIALTY SECTION This article was submitted to Plant Pathogen Interactions, a section of the journal Frontiers in Plant Science RECEIVED 22 October 2022 ACCEPTED 21 December 2022 PUBLISHED 23 January 2023 CITATION Zhang C, Lane B, Fern´andez-Campos M, Cruz-Sancan A, Lee D-Y, Gongora-Canul C, Ross TJ, Da Silva CR, Telenko DEP, Goodwin SB, Scoﬁeld SR, Oh S, Jung Jand Cruz CD (2023) Monitoring tar spot disease in corn at different canopy and temporal levels using aerial multispectral imaging and machine learning. Front. Plant Sci. 13:1077403 . doi: 10.3389/fpls.2022.1077403 COPYRIGHT © 2023 Zhang, Lane, Fern´andez-Campos, Cruz-Sancan, Lee, Gongora-Canul, Ross, Da Silva, Telenko, Goodwin, Scoﬁeld, Oh, Jung and Cruz. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Monitoring tar spot disease in corn at different canopy and temporal levels using aerial multispectral imaging and machine learning Chongyuan Zhang 1, Brenden Lane 1, Mariela Fern´andez-Campos 1, Andres Cruz-Sancan 1, Da-Young Lee 1, Carlos Gongora-Canul 1,2, Tiffanna J. Ross 1, Camila R. Da Silva 1, Darcy E. P. Telenko 1, Stephen B. Goodwin 3, Steven R. Scoﬁeld3, Sungchan Oh 4, Jinha Jung 5 and C. D. Cruz 1* 1 Department of Botany and Plant Pathology, Purdue University, West Lafayette, IN, United States, 2Tecnolo´ gico Nacional de Me´xico, Instituto Tecnolo´gico de Conkal, Yucat´an, Mexico, 3 USDAAgricultural Research Service, Crop Production and Pest Control Research Unit, West Lafayette, IN, United States, 4 Institute for Plant Sciences, Purdue University, West Lafayette, IN, United States, 5 Lyles School of Civil Engineering, Purdue University, West Lafayette, IN, United States Introduction: Tar spot is a high-proﬁle disease, causing various degrees of yield losses on corn (Zea mays L.) in several countries throughout the Americas. Disease symptoms usually appear at the lower canopy in corn ﬁelds with a history of tar spot infection, making it difﬁcult to monitor the disease with unmanned aircraft systems (UAS) because of occlusion. Methods: UAS-based multispectral imaging and machine learning were used to monitor tar spot at different canopy and temporal levels and extract epidemiological parameters from multiple treatments. Disease severity was assessed visually at three canopy levels within micro-plots, while aerial images were gathered by UASs equipped with multispectral cameras. Both disease severity and multispectral images were collected from ﬁve to eleven time points each year for two years. Image-based features, such as single-band reﬂectance, vegetation indices (VIs), and their statistics, were extracted from ortho-mosaic images and used as inputs for machine learning to develop disease quantiﬁcation models. Results and discussion: The developed models showed encouraging performance in estimating disease severity at different canopy levels in both years (coefﬁcient of determination up to 0 . 93 and Lin ’ s concordance Frontiers in Plant Science 01 [frontiersin.org](frontiersin.org) correlation coefﬁcient up to 0 . 97) . Epidemiological parameters, including initial disease severity or y0 and area under the disease progress curve, were modeled using data derived from m

FAQs

How was tar spot disease monitored in the study?

The study used unmanned aircraft systems (UAS) with multispectral cameras to collect aerial imagery, alongside visual assessments of disease severity at different canopy levels in micro-plots.

What data were used to train the machine learning disease quantification models?

Models used image-based features extracted from ortho-mosaic multispectral images, including single-band reflectance, vegetation indices, and their statistical summaries.

What were the key findings about model performance and practical use?

The developed models showed encouraging performance for estimating disease severity across canopy levels in both years, and digital phenotyping enabled monitoring of disease onset even when severity was relatively low (below 1%).

Aug 5, 2026

Upload by Jiven