Recognition method for tobacco leaves curing stages in step-type curing barns using improved YOLOv8

doi:10.14088/j.cnki.issn0439-8114.2025.09.032

Abstract

Abstract: To address the need for recognizing tobacco leaves curing stages in step-type curing barns, a real-time detection model named T-YOLOv8 based on improved YOLOv8 was proposed. First, a lightweight EfficientViT was introduced as the backbone network to enhance the detection accuracy and inference speed of the model in highly complex scenarios. Second, by designing a deformable adaptive attention mechanism module (DA-Attention), the model’s feature fusion and expression capabilities under different environmental conditions were enhanced, further improving its robustness with diverse data inputs.Finally, a cross-stage feature extraction module (CSPStage) and an improved loss function (Focal EIoU_loss) were incorporated to optimize the efficiency of target feature extraction and fusion while reducing the computational cost of the model. The results showed that the precision, recall, and mean average precision (mAP50) of the T-YOLOv8 model reached 90.3%, 94.1%, and 95.2%, respectively, representing improvements of 5.2%, 4.1%, and 5.2% compared to the YOLOv8 model. Compared to the YOLO series models and classical object detection models, the T-YOLOv8 model demonstrated significant advantages in real-time performance and accuracy. The T-YOLOv8 model achieved real-time monitoring of tobacco leaves curing status in step-type curing barns, providing support for the construction of intelligent and automated curing systems.

Key words: improved YOLOv8, step-type curing barns, tobacco leaves curing, recognition method

CLC Number:

TP391

WANG Chun-qing, SHANG Shu-qi, ZHANG Xi-ya, LIU Wei, YUE Dan-song. Recognition method for tobacco leaves curing stages in step-type curing barns using improved YOLOv8[J]. HUBEI AGRICULTURAL SCIENCES, 2025, 64(9): 202-212.

References 28

[1]	黄宁,祖庆学,刘红峰,等.空气源热泵烤房与生物质新能源烤房对烤后烟叶经济效益的影响[J].贵州农业科学,2020,48(3):141-143.
[2]	李静浩,孙光伟,陈振国,等.烟叶烘烤技术研究进展与智能烘烤技术展望[J].湖南文理学院学报(自然科学版),2021,33(4):88-92.
[3]	白万明,黄逸兰,徐希祥,等.密集烤房物联网智能烤烟系统开发与应用研究[J].浙江农业科学,2022,63(7):1610-1616.
[4]	THOMAS C E.Techniques of image analysis applied to the measurement of tobacco and related products[A] 42nd Tobacco Chemists’ Research Conference[C].North Carolina StateUniversity:Tobacco science research conference papers, 1988.
[5]	ZHANG J, SOKHANSANJ S, WU S, et al.A trainable grading system for tobacco leaves[J]. Computers and electronics in agriculture, 1997, 16(3): 231-244.
[6]	YAWOOTTI A, KAEWTRAKULPONG P.A machine vision system for Thai flue-cured tobacco classification[A].Electrical engineering/electronics, computer, telecommunications, and information technology international conference[C].Thailand: IEEE Xplore,2005.
[7]	HAN L.Recognition of the part of growth of flue-cured tobacco leaves based on support vector machine[A].Proceedings of the 2008 7th world congress on intelligent control and automation[C]. New York:IEEE, 2008.
[8]	汪睿琪,张炳辉,顾钢,等.基于YOLOv5的鲜烟叶成熟度识别模型研究[J].中国烟草学报,2023,29(2):46-55.
[9]	LI Q, SHAO Z, ZHOU W, et al.MobileOne-YOLO: Improving the YOLOv7 network for the detection of unfertilized duck eggs and early duck embryo development-a novel approach[J]. Computers and electronics in agriculture, 2023, 214: 108316.
[10]	WU X, WU X, LI D, et al.Tobacco leaves maturity classification based on deep learning and proximal hyperspectral imaging[J]. Analytical letters, 2024, 57(13): 2034-2049.
[11]	LI J X, ZHAO H, ZHU S P, et al.An improved lightweight network architecture for identifying tobacco leaf maturity based on deep learning[J]. Journal of intelligent and fuzzy systems, 2021, 41(2): 4149-4158.
[12]	张成双,王先伟,刘志刚,等.基于深度学习的烟叶烘烤实时识别研究[J].智慧农业导刊,2022,2(21):18-22.
[13]	姜增昀,朱世平,冯川,等.高精度轻量级烟叶烘烤阶段识别模型研究[J].中国烟草学报,2023,29(1):55-63.
[14]	Ultralytics[EB/OL].(2023-01-10)[2023-05-20]. https://github.com/ultralytics/ultralytics.
[15]	CAI H, LI J, HU M, et al. EfficientViT: Multi-scale linear attention for high-resolution dense prediction[J]. arXiv preprint arXiv:2205.14756, 2022.
[16]	SANDLER M, HOWARD A, ZHU M, et al.MobileNetV2: Inverted residuals and linear bottlenecks[A].Proceedings of the IEEE conference on computer vision and pattern recognition[C].Salt palace convention center:IEEE computer society, 2018.
[17]	XIA Z, PAN X, SONG S, et al.Vision Transformer with deformable attention[A].Proceedings of the IEEE/CVF conference on computer vision and pattern recognition[C].New Orleans, LA, USA:IEEE, 2022.4794-4803.
[18]	WANG C Y, BOCHKOVSKIY A, LIAO H Y M. Scaled-YOLOv4: Scaling cross stage partial network[A].Proceedings of the IEEE/CVF conference on computer vision and pattern recognition[C].Virtual conference:IEEE,2021.13029-13038.
[19]	ZHENG Z, WANG P, REN D, et al.Enhancing geometric factors in model learning and inference for object detection and instance segmentation[J]. IEEE transactions on cybernetics,2021,52(8): 8574-8586.
[20]	ZHANG Y F, REN W, ZHANG Z, et al.Focal and efficient IoU loss for accurate bounding box regression[J]. Neurocomputing, 2022, 506: 146-157.
[21]	AL-FURAIJI O J, ANH TUAN N, TSVIATKOU V Y. A new fast efficient non-maximum suppression algorithm based on image segmentation[J]. Indonesian journal of electrical engineering and computer science, 2020, 19(2): 1062-1070.
[22]	WANG C Y, YEH I H, MARK LIAO H Y. Yolov9: Learning what you want to learn using programmable gradient information[A].European conference on computer vision[C]. Cham: Springer nature switzerland, 2024.
[23]	CHEN P Y, ZHOU J, ZHANG S, et al.Parallel residual Bi-fusion feature pyramid network for accurate single-shot object detection[J]. IEEE transactions on image processing, 2021, 30: 9099-9111.
[24]	LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[A].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition[C].Hawaii convention center: IEEE computer society,2017.2117-2125.
[25]	CHEN Y M, JIANG H, ZHOU Y Q, et al. YOLO-MS: Rethinking multi-scale representation learning for real-time object detection[J]. arXiv preprint arXiv:2308.05480, 2023.
[26]	SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM: Why did you say that?[J]. arXiv preprint arXiv:1611.07450, 2016.
[27]	李增盛,孟令峰,王松峰,等.基于图像处理的烟叶烘烤阶段判别模型优选[J].中国烟草学报,2022,28(2):65-76.
[28]	邢玉清,樊彩霞,豆根生,等.基于小波核极限学习机的烟叶烘烤过程的智能识别[J].中国烟草学报,2024,30(1):55-62.