Abstract: Tea diseases, including brown and gray blight, result in significant yield and quality losses, especially in Longjing tea production. Traditional detection methods are prone to errors, while existing deep learning models often struggle to be robust under natural field conditions. To address these challenges, an improved lightweight detection model, asymmetric multi-level (AML) mechanism, dynamic snake convolution (DSC), and scalable intersection over union (SIoU) loss function-You Only Look Once (YOLO) (ADS-YOLO), was developed and validated. In the method, a dataset comprising 5694 smartphone-captured images of tea leaves was established under natural lighting. Enhancements were implemented in the YOLO11n baseline algorithm through incorporation of the SIoU loss function for better bounding box regression, DSC, which realizes adaptive feature extraction based on the dynamic spatial context, and an AML mechanism, which achieves lightweight feature fusion via adaptive multi-scale design. The results showed that ADS-YOLO achieved a precision of 0.935 and a recall of 0.870, compared to 0.894 and 0.818, respectively, when the baseline YOLO11n was used. Importantly, ADS-YOLO demonstrated a real-time performance of 137.1 frames per second (FPS), coupled with reduced computational costs. ADS-YOLO improved the mean average precision (mAP) at intersection over union threshold of 0.5 (mAP@0.5) by 6.4% compared with YOLOv5n and achieved up to 44.6% higher accuracy than YOLOv7t. In conclusion, ADS-YOLO achieved high accuracy, providing a scalable solution for real-time crop health monitoring and sustainable precision agriculture for tea production.

在YOLO11n算法中改进RGB图像识别以实现对茶树病害的精确检测

[1]AbuliziA, YeJX, AbudukelimuH, et al., 2025. DM-YOLO: improved YOLOv9 model for tomato leaf disease detection. Front Plant Sci, 15:1473928.

[2]AnQL, WangK, LiZY, et al., 2022. Real-time monitoring method of strawberry fruit growth state based on YOLO improved model. IEEE Access, 10:124363-124372.

[3]AriyawansaHA, TsaiI, WangJY, et al., 2021. Molecular phylogenetic diversity and biological characterization of Diaporthe species associated with leaf spots of Camellia sinensis in Taiwan. Plants, 10(7):1434.

[4]BaoWX, FanT, HuGS, et al., 2022. Detection and identification of tea leaf diseases based on AX-RetinaNet. Sci Rep, 12:2183.

[5]ChenHB, WangRJ, DuJM, et al., 2023. Feature refinement method based on the two-stage detection framework for similar pest detection in the field. Insects, 14(10):819.

[6]ChutichaimaytarP, ZongqiZ, KaewtrakulpongK, et al., 2025. An improved small object detection CTB-YOLO model for early detection of tip-burn and powdery mildew symptoms in coriander (Coriandrum sativum) for indoor environment using an edge device. Smart Agric Technol, 12:101142.

[7]DaiM, DorjoyMMH, MiaoH, et al., 2023. A new pest detection method based on improved YOLOv5m. Insects, 14:54.

[8]DhanyaVG, SubeeshA, KushwahaNL, et al., 2022. Deep learning based computer vision approaches for smart agricultural applications. Artif Intell Agric, 6:211-229.

[9]DongC, DuoqianM, 2023. Control distance IoU and control distance IoU loss for better bounding box regression. Pattern Recognit, 137:109256.

[10]ElharroussO, AkbariY, AlmadeedN, et al., 2024. Backbones-review: feature extractor networks for deep learning and deep reinforcement learning approaches in computer vision. Comput Sci Rev, 53:100645.

[11]FengHR, ChenXQ, DuanZY, 2025. LCDDN-YOLO: lightweight cotton disease detection in natural environment, based on improved YOLOv8. Agriculture, 15(4):421.

[12]GuZC, ZhuK, YouST, 2023. YOLO-SSFS: a method combining SPD-Conv/STDL/IM-FPN/SIoU for outdoor small target vehicle detection. Electronics, 12(18):3744.

[13]HuangZX, GoudaM, YeST, et al., 2024. Advanced deep learning algorithm for instant discriminating of tea leave stress symptoms by smartphone-based detection. Plant Physiol Biochem, 212:108769.

[14]IqbalMS, El-AshramS, HussainS, et al., 2019. Efficient cell classification of mitochondrial images by using deep learning. J Opt, 48:113-122.

[15]LawalOM, HuaminZ, FanZ, 2021. Ablation studies on YOLOFruit detection algorithm for fruit harvesting robot using deep learning. IOP Conf Ser Earth Environ Sci, 922:012001.

[16]LiXT, ZhangTH, YuM, et al., 2025. A YOLOv8‐based method for detecting tea disease in natural environments. Agron J, 117(2):e70043.

[17]LinYT, XiaYJ, XiaPC, et al., 2025. YOLO11-ARAF: an accurate and lightweight method for apple detection in real-world complex orchard environments. Agriculture, 15(10):1104.

[18]LiuY, LiuHM, XuWH, et al., 2024. Advances and challenges of carbon storage estimation in tea plantation. Ecol Inf, 81:102616.

[19]LuJQ, LuoHX, YuCR, et al., 2024. Tea bud DG: a lightweight tea bud detection model based on dynamic detection head and adaptive loss function. Comput Electron Agric, 227:109522.

[20]MaCR, ChiG, JuXP, et al., 2025. YOLO-CWD: a novel model for crop and weed detection based on improved YOLOv8. Crop Prot, 192:107169.

[21]MengJQ, WangYX, ZhangJM, et al., 2023. Tea bud and picking point detection based on deep learning. Forests, 14(6):1188.

[22]ParasharN, JohriP, ElbeltagiA, et al., 2025. Enhanced residual-attention deep neural network for disease classification in maize leaf images. Sci Rep, 15:29452.

[23]PerezL, WangJ, 2017. The effectiveness of data augmentation in image classification using deep learning. arXiv:1712.04621.

[24]RahatIS, GhoshH, DaraS, et al., 2025. Towards precision agriculture tea leaf disease detection using CNNs and image processing. Sci Rep, 15:17571.

[25]SoebJA, JubayerF, TarinTA, et al., 2023. Tea leaf disease detection and identification based on YOLOv7 (YOLO-T). Sci Rep, 13:6078.

[26]SunYG, LiZH, GuoHP, et al., 2025. TDDet: a novel lightweight and efficient tea disease detector. Comput Electron Agric, 237:110481.

[27]WangRJ, ChenYS, ZhangGH, et al., 2025. YOLO11-PGM: high-precision lightweight pomegranate growth monitoring model for smart agriculture. Agronomy, 15(5):1123.

[28]WangX, WuY, CuiLF, et al., 2025. Linear pattern detection of building groups by integrating dynamic snake convolution with YOLO11. Geocarto Int, 40(1):2471914.

[29]WangZH, ZhaoZX, HongN, et al., 2017. Characterization of causal agents of a novel disease inducing brown-black spots on tender tea leaves in China. Plant Dis, 101(10):1802-1811.

[30]WenCM, ChengY, LiSP, et al., 2025. Slim-YOLO: an improved sugarcane tail tip recognition algorithm based on YOLO11n for complex field environments. Appl Sci, 15(8):4286.

[31]WuX, DengHY, WangQ, et al., 2023. Meta‐learning shows great potential in plant disease recognition under few available samples. Plant J, 114(4):767-782.

[32]XieS, SunHW, 2023. Tea-YOLOv8s: a tea bud detection model based on deep learning and computer vision. Sensors, 23(14):6576.

[33]XuJL, PanF, HanXH, et al., 2024. EdgeTrim-YOLO: improved trim YOLO framework tailored for deployment on edge devices. 2024 4th International Conference on Computer Communication and Artificial Intelligence (CCAI). Xi’an, China, p.113-118.

[34]YangZM, MaW, LuJZ, et al., 2023. The application status and trends of machine vision in tea production. Appl Sci, 13(19):10744.

[35]YuanZH, NingH, TangXY, et al., 2024. GDCP-YOLO: enhancing steel surface defect detection using lightweight machine learning approach. Electronics, 13(7):1388.

[36]ZhanBS, XiongX, LiXL, et al., 2024. BHC-YOLOV8: improved YOLOv8-based BHC target detection model for tea leaf disease and defect in real-world scenarios. Front Plant Sci, 15:1492504.

[37]ZhengZH, WangP, LiuW, et al., 2020. Distance-IoU loss: faster and better learning for bounding box regression. Proceedings of the AAAI Conference on Artificial Intelligence. AAAI Press, Washington, DC, USA, 34(7):12993-13000.

[38]ZhengZW, YuWW, 2025. RG-YOLO: multi-scale feature learning for underwater target detection. Multimedia Syst, 31:26.

[39]ZhuJJ, QiuJL, ChenSW, et al., 2025. An application of YOLOv8 integrated with attention mechanisms for detection of grape leaf black rot spots. PLoS One, 20(4):e0321788.

[40]ZhuXY, ChenFJ, ZhengYL, et al., 2024. Detection of Camellia oleifera fruit maturity in orchards based on modified lightweight YOLO. Comput Electron Agric, 226:109471.

[41]ZongHT, ZhangYS, LiuXX, et al., 2023. Recent trends in smartphone‐based optical imaging biosensors for genetic testing: a review. View, 4(4):20220062.

Similar articles