Early detection and precise management of tea leaf diseases and pests are essential for maintaining tea yield and quality. In China, the average annual loss attributed to tea leaf diseases and pests is approximately 15%-20%. Conventional manual inspection is costly, time-consuming, and poorly suited to rapid early warning under complex tea plantation conditions. To address these limitations, this study proposes and implements a real-time tea leaf disease detection system that integrates YOLOv10 with a customized handheld terminal. For hardware-based data acquisition, a utility-model-patented touch-controlled handheld imaging terminal was developed. The terminal incorporates a rotary stage and an anti-slip spring-seat mechanism, which effectively suppresses device shake during one-handed field operation and improves the stability of image acquisition. For algorithmic recognition, an efficient YOLOv10 object detector with a non-maximum suppression (NMS)-free design is introduced and deployed on an edge computing node. Experimental results on a test set containing healthy leaves and seven representative diseases (algal leaf disease, anthracnose, bird's-eye spot, brown blight, gray spot disease, red leaf spot disease, and white spot disease) show that the proposed system achieves an overall mean average precision ([email protected]) of 0.753. In particular, the recognition precision for visually distinctive diseases such as red leaf spot disease (redleafspot) reaches 0.950. The system also achieves an inference speed of 150 FPS, satisfying the low-latency requirements of real-time detection. By coupling customized hardware with a lightweight deep learning model, the proposed system reduces detection cost and response time, providing a practical engineering solution for disease monitoring and automation in smart tea plantations.

Xiao Li, Bin Yang, Jiayi Luo, Guixiu Xie, Xiaoliang Liang, Jiaxue Li. A Real-Time Tea Leaf Disease Detection System for Tea Leaves Based on YOLOv10 and a Customized Handheld Terminal. International Journal of Frontiers in Engineering Technology (2026), Vol. 8, Issue 2: 45-54. https://doi.org/10.25236/IJFET.2026.080207.

[1] Wu H C. Integrated Control Techniques for Brown Leaf Spot Disease of Tea [J]. Agriculture and Technology, 2015(16). DOI:10.11974/nyyjs.20150833093.

[2] Wang K, Liu D M. Intelligent Recognition Method for Tea Leaf States Based on Deep Learning [J]. Journal of Chongqing University of Technology (Natural Science), 2015(12). DOI:10.3969/j.issn.1674-8425(z).2015.12.020.

[3] Chen L J, Xu B L. Discussion on Approaches to Improving Tea Output Value [J]. Guizhou Agricultural Sciences, 2008(3). DOI:10.3969/j.issn.1001-3601.2008.03.071.

[4] Zou Y, Hu G G. Tea Plucking and Management [J]. Anhui Agricultural Science Bulletin, 2005, 11(1): 71. DOI:10.3969/j.issn.1007-7731.2005.01.061.

[5] Wang A, Chen H, Liu L, Chen K, Lin Z, Han J, et al. YOLOv10: Real-Time End-to-End Object Detection [J]. arXiv preprint arXiv:2405.14458, 2024. DOI:10.48550/arXiv.2405.14458.

[6] Guo L, Chen H B, Cui Z M, Lai Z H, Huang W L, Lin T L, et al. GDE-YOLO: a robust and accurate method for real-time tea leaf disease detection in complex plantation environments [J]. Engineering Agriculture, 2026, 13(4): 26668-26681. DOI:10.15302/J-FASE-2026669.

[7] Sun Y G, Wu F, Zhou Q Y. A deep learning detection algorithm for tea diseases [J]. Journal of Xinyang Normal University (Natural Science Edition), 2024, 37(2): 246-251.

[8] Ayon R, Mubin A S, Hridoy A A G, Mojumdar M U, Chakraborty N R, Islam M. Deep Learning for Tea Leaf Disease Detection: Comparing VGG16, Vision Transformer, EfficientNetB3, and Xception [C]//2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT). IEEE, 2024: 1-6. DOI:10.1109/ICCCNT61001.2024.10726203.

[9] Huang F, Liu F, Wang Y, Luo F. Research Progress and Prospect on Computer Vision Technology Application in Tea Production [J]. Journal of Tea Science, 2019, 39(1): 81-87. DOI:10.13305/j.cnki.jts.2019.01.009.

[10] Chen J, Zhang D, Nanehkaran Y A, Li D. Identification of plant disease images via a CNN-based region proposal approach [J]. IET Image Processing, 2020, 14(11): 2384-2393.

[11] Ge Z, Liu S, Wang F, Li Z, Sun J. YOLOX: Exceeding YOLO Series in 2021 [J]. arXiv preprint arXiv:2107.08430, 2021.

[12] Wang Y X, Zhang Y, Yang C Y, et al. Research Progress in Image Recognition Technologies for Crop Diseases Based on Deep Learning [J]. Acta Agriculturae Zhejiangensis, 2019(4). DOI:10.3969/j.issn.1004-1524.2019.04.21.

[13] Huang X Y, Li J, Jiang S. Research on Rice Variety Recognition Technology Based on Computer Vision [J]. Journal of Jiangsu University (Natural Science Edition), 2004(2). DOI:10.3969/j.issn.1671-7775.2004.02.003.

[14] Long D F, Fan S C. Application of Computer Vision to Log Volume Measurement [J]. Chinese Journal of Scientific Instrument, 2004(z1). DOI:10.3321/j.issn:0254-3087.2004.z1.432.

[15] Lan J H, Wang D, Shen X P. Research Progress of Convolutional Neural Networks in Visual Image Detection [J]. Chinese Journal of Scientific Instrument, 2020(4). DOI:10.19650/j.cnki.cjsi.J2006003.

[16] Sun X X. The Research of Tea Buds Detection and Leaf Diseases Image Recognition Based on Deep Learning [D]. Tai'an: Shandong Agricultural University, 2019.

[17] Chen M T. Recognition and Location of High-Quality Tea Buds Based on Computer Vision [D]. Qingdao: Qingdao University of Science and Technology, 2019.