LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease

Abstract Quick disease diagnosis is crucial in agriculture to reduce crop yield losses. Deep learning techniques have already been successfully applied to several tasks in agriculture, but it requires sufficient datasets. Obtaining sufficient plant disease images is challenging due to seasonality an...

Full description

Saved in:
Bibliographic Details
Main Authors: Ri Zheng, Jimin Lee, Helin Yin, Dong Jin, Yeong Hyeon Gu
Format: Article
Language:English
Published: Springer 2025-06-01
Series:Journal of King Saud University: Computer and Information Sciences
Subjects:
Online Access:https://doi.org/10.1007/s44443-025-00070-x
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849235369628794880
author Ri Zheng
Jimin Lee
Helin Yin
Dong Jin
Yeong Hyeon Gu
author_facet Ri Zheng
Jimin Lee
Helin Yin
Dong Jin
Yeong Hyeon Gu
author_sort Ri Zheng
collection DOAJ
description Abstract Quick disease diagnosis is crucial in agriculture to reduce crop yield losses. Deep learning techniques have already been successfully applied to several tasks in agriculture, but it requires sufficient datasets. Obtaining sufficient plant disease images is challenging due to seasonality and a shortage of experts. Generative Adversarial Networks (GANs) can address this issue by generating synthetic images, but existing methods struggle with complex diseases like fire blight, where symptoms are localized. To overcome this problem, we propose LITP-GAN, a novel image-to-image translation pipeline considering the lesion area. LITP-GAN includes three stages: (Stage 1) Lesion area detection using DEtection TRansformer (DETR) to detect lesion areas, (Stage 2) Image translation via Global and Local Alignment Networks (GLA-Net) to generate diseased leaf images, and (Stage 3) Post-Processing using Seamless Cloning method for generating realistic images by minimizing visual inconsistencies. We tested 400 healthy leaf images, and 236 generated images were identified as realistic diseased images. Applying these to a real field plant disease detection model showed significant improvements: 7.57% in precision, 4% in recall, and 3.5% in mAP compared to models trained on the original datasets. These findings suggest that LITP-GAN-augmented datasets can improve real field plant disease diagnosis, providing a promising tool for agricultural applications.
format Article
id doaj-art-1c0d13d709834a0f9497a5c9932f5b50
institution Kabale University
issn 1319-1578
2213-1248
language English
publishDate 2025-06-01
publisher Springer
record_format Article
series Journal of King Saud University: Computer and Information Sciences
spelling doaj-art-1c0d13d709834a0f9497a5c9932f5b502025-08-20T04:02:49ZengSpringerJournal of King Saud University: Computer and Information Sciences1319-15782213-12482025-06-0137412110.1007/s44443-025-00070-xLITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant diseaseRi Zheng0Jimin Lee1Helin Yin2Dong Jin3Yeong Hyeon Gu4Department of Artificial Intelligence and Data Science, College of AI Convergence, Sejong UniversityDepartment of Artificial Intelligence and Data Science, College of AI Convergence, Sejong UniversityDepartment of Artificial Intelligence and Data Science, College of AI Convergence, Sejong UniversityDepartment of Artificial Intelligence and Data Science, College of AI Convergence, Sejong UniversityDepartment of Artificial Intelligence and Data Science, College of AI Convergence, Sejong UniversityAbstract Quick disease diagnosis is crucial in agriculture to reduce crop yield losses. Deep learning techniques have already been successfully applied to several tasks in agriculture, but it requires sufficient datasets. Obtaining sufficient plant disease images is challenging due to seasonality and a shortage of experts. Generative Adversarial Networks (GANs) can address this issue by generating synthetic images, but existing methods struggle with complex diseases like fire blight, where symptoms are localized. To overcome this problem, we propose LITP-GAN, a novel image-to-image translation pipeline considering the lesion area. LITP-GAN includes three stages: (Stage 1) Lesion area detection using DEtection TRansformer (DETR) to detect lesion areas, (Stage 2) Image translation via Global and Local Alignment Networks (GLA-Net) to generate diseased leaf images, and (Stage 3) Post-Processing using Seamless Cloning method for generating realistic images by minimizing visual inconsistencies. We tested 400 healthy leaf images, and 236 generated images were identified as realistic diseased images. Applying these to a real field plant disease detection model showed significant improvements: 7.57% in precision, 4% in recall, and 3.5% in mAP compared to models trained on the original datasets. These findings suggest that LITP-GAN-augmented datasets can improve real field plant disease diagnosis, providing a promising tool for agricultural applications.https://doi.org/10.1007/s44443-025-00070-xComputer visionData augmentationDeep learningImage-to-image translationPlant disease detection
spellingShingle Ri Zheng
Jimin Lee
Helin Yin
Dong Jin
Yeong Hyeon Gu
LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
Journal of King Saud University: Computer and Information Sciences
Computer vision
Data augmentation
Deep learning
Image-to-image translation
Plant disease detection
title LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
title_full LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
title_fullStr LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
title_full_unstemmed LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
title_short LITP-GAN: Leaf image translation pipeline using DETR and GLA-Net considering the lesion area of plant disease
title_sort litp gan leaf image translation pipeline using detr and gla net considering the lesion area of plant disease
topic Computer vision
Data augmentation
Deep learning
Image-to-image translation
Plant disease detection
url https://doi.org/10.1007/s44443-025-00070-x
work_keys_str_mv AT rizheng litpganleafimagetranslationpipelineusingdetrandglanetconsideringthelesionareaofplantdisease
AT jiminlee litpganleafimagetranslationpipelineusingdetrandglanetconsideringthelesionareaofplantdisease
AT helinyin litpganleafimagetranslationpipelineusingdetrandglanetconsideringthelesionareaofplantdisease
AT dongjin litpganleafimagetranslationpipelineusingdetrandglanetconsideringthelesionareaofplantdisease
AT yeonghyeongu litpganleafimagetranslationpipelineusingdetrandglanetconsideringthelesionareaofplantdisease