Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models

Powdery mildew is one of the most common diseases affecting strawberry yield and quality. Accurate and timely detection is essential to reduce pesticide usage and labor costs. However, recognizing strawberry powdery mildew in complex field environments remains a significant challenge. In this study,...

Full description

Saved in:
Bibliographic Details
Main Authors: Jingzhi Wang, Jiayuan Li, Fanjia Meng
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:AgriEngineering
Subjects:
Online Access:https://www.mdpi.com/2624-7402/7/6/182
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849472679895105536
author Jingzhi Wang
Jiayuan Li
Fanjia Meng
author_facet Jingzhi Wang
Jiayuan Li
Fanjia Meng
author_sort Jingzhi Wang
collection DOAJ
description Powdery mildew is one of the most common diseases affecting strawberry yield and quality. Accurate and timely detection is essential to reduce pesticide usage and labor costs. However, recognizing strawberry powdery mildew in complex field environments remains a significant challenge. In this study, an HSV-based image segmentation method was employed to enhance the extraction of disease regions from complex backgrounds. A total of 14 widely used deep learning models—including SqueezeNet, GoogLeNet, ResNet-50, AlexNet, and others—were systematically evaluated for their classification performance. To address sample imbalance, data augmentation was applied to 2372 healthy and 553 diseased leaf images, resulting in 11,860 training samples. Experimental results showed that InceptionV4, DenseNet-121, and ResNet-50 achieved superior performance across metrics such as accuracy, F1-score, recall, and loss. Models such as MobileNetV2, AlexNet, VGG-16, and InceptionV3 demonstrated certain strengths, and models like SqueezeNet, VGG-19, EfficientNet, and even ResNet-50 showed room for further improvement in performance. These findings demonstrate that CNN models originally developed for other crop diseases can be effectively adapted to detect strawberry powdery mildew under complex conditions. Future work will focus on enhancing model robustness and deploying the system for real-time field monitoring.
format Article
id doaj-art-fb3b8d3cbfdf4f9f92d1e8609da9dfe8
institution Kabale University
issn 2624-7402
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series AgriEngineering
spelling doaj-art-fb3b8d3cbfdf4f9f92d1e8609da9dfe82025-08-20T03:24:28ZengMDPI AGAgriEngineering2624-74022025-06-017618210.3390/agriengineering7060182Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning ModelsJingzhi Wang0Jiayuan Li1Fanjia Meng2College of Information and Electrical Engineering, China Agricultural University, No. 17, Qinghua East Road, Haidian District, Beijing 100091, ChinaCollege of Information and Electrical Engineering, China Agricultural University, No. 17, Qinghua East Road, Haidian District, Beijing 100091, ChinaCollege of Information and Electrical Engineering, China Agricultural University, No. 17, Qinghua East Road, Haidian District, Beijing 100091, ChinaPowdery mildew is one of the most common diseases affecting strawberry yield and quality. Accurate and timely detection is essential to reduce pesticide usage and labor costs. However, recognizing strawberry powdery mildew in complex field environments remains a significant challenge. In this study, an HSV-based image segmentation method was employed to enhance the extraction of disease regions from complex backgrounds. A total of 14 widely used deep learning models—including SqueezeNet, GoogLeNet, ResNet-50, AlexNet, and others—were systematically evaluated for their classification performance. To address sample imbalance, data augmentation was applied to 2372 healthy and 553 diseased leaf images, resulting in 11,860 training samples. Experimental results showed that InceptionV4, DenseNet-121, and ResNet-50 achieved superior performance across metrics such as accuracy, F1-score, recall, and loss. Models such as MobileNetV2, AlexNet, VGG-16, and InceptionV3 demonstrated certain strengths, and models like SqueezeNet, VGG-19, EfficientNet, and even ResNet-50 showed room for further improvement in performance. These findings demonstrate that CNN models originally developed for other crop diseases can be effectively adapted to detect strawberry powdery mildew under complex conditions. Future work will focus on enhancing model robustness and deploying the system for real-time field monitoring.https://www.mdpi.com/2624-7402/7/6/182strawberry powdery mildewdeep learningconvolutional neural networkdisease detection
spellingShingle Jingzhi Wang
Jiayuan Li
Fanjia Meng
Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
AgriEngineering
strawberry powdery mildew
deep learning
convolutional neural network
disease detection
title Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
title_full Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
title_fullStr Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
title_full_unstemmed Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
title_short Recognition of Strawberry Powdery Mildew in Complex Backgrounds: A Comparative Study of Deep Learning Models
title_sort recognition of strawberry powdery mildew in complex backgrounds a comparative study of deep learning models
topic strawberry powdery mildew
deep learning
convolutional neural network
disease detection
url https://www.mdpi.com/2624-7402/7/6/182
work_keys_str_mv AT jingzhiwang recognitionofstrawberrypowderymildewincomplexbackgroundsacomparativestudyofdeeplearningmodels
AT jiayuanli recognitionofstrawberrypowderymildewincomplexbackgroundsacomparativestudyofdeeplearningmodels
AT fanjiameng recognitionofstrawberrypowderymildewincomplexbackgroundsacomparativestudyofdeeplearningmodels