ORIGINAL RESEARCH article
Front. Plant Sci.
Sec. Sustainable and Intelligent Phytoprotection
TDS-YOLO: A Lightweight Detection Model for Fine-Grained Segmentation of Tea Leaf Diseases
1. Universiti Putra Malaysia, Serdang, Malaysia
2. Huzhou University, Huzhou, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Abstract
Abstract: Timely identification and precise segmentation of tea leaf diseases are essential for intelligent agricultural management. However, achieving a balance between lightweight deployment and high-precision segmentation remains challenging due to uneven illumination, background interference, and the subtle textures of early-stage lesions in natural environments. To address these issues, we propose TDS-YOLO, a lightweight segmentation model based on the YOLOv11 framework. The model incorporates three primary innovations: the C3K2_EViM_CGLU module for global dependency modeling, the EfficientHead for lightweight pixel-level representation, and the C2PSA_Mona module to enhance multi-scale texture perception. Experimental results on a diverse dataset of 4,933 images show that TDS-YOLO achieves state-of-the-art performance with only 2.53M parameters. It attains an mAP@0.5 of 90.1% for both detection and segmentation, significantly outperforming YOLOv11-seg and other mainstream models while maintaining an inference speed of 96 FPS. This work provides a highly efficient and robust solution for real-time monitoring of tea diseases, offering a technical breakthrough for precision tea plantation management and the broader field of smart digital agriculture.
Summary
Keywords
deep learning, EfficientHead, TDS-YOLO, Tea Leaf Disease Segmentation, YOLOv11
Received
16 December 2025
Accepted
18 February 2026
Copyright
© 2026 Xie, Wang, Zu, Jusoh and Jia. 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) or licensor 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.
*Correspondence: Yusmadi Jusoh; Liangquan Jia
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.