基于改进YOLOv4-tiny的茶叶嫩芽检测模型

doi:10.13305/j.cnki.jts.2022.04.009

摘要/Abstract

摘要： 精准检测茶叶嫩芽是茶叶机械智能采摘的重要前提。针对茶叶大小不一、遮挡造成的小尺度嫩芽特征显著性弱、漏检率高等问题,提出一种基于改进YOLOv4-tiny的茶叶嫩芽检测模型。该模型在颈部网络添加52×52的浅层特征层以提高YOLOv4-tiny网络对小目标嫩芽的关注度,通过引入卷积块注意力机制（Convolutional block attention module,CBAM）以抑制背景噪声,提高嫩芽特征的显著性,采用双向特征金字塔网络（Bidirectional feature pyramid network,BiFPN）以融合不同尺度的特征信息,从而提出一个高性能轻量化的茶叶嫩芽检测模型YOLOv4-tiny-Tea。对同一训练集与测试集进行模型训练与性能测试,结果表明YOLOv4-tiny-Tea模型检测精确率和召回率分别为97.77%和95.23%,相比改进之前分别提高了5.58个百分点和23.14个百分点。消融试验验证了网络结构改进对不同尺度嫩芽检测的有效性,并将改进后的YOLOv4-tiny-Tea模型与3种YOLO系列算法进行对比,发现改进后的YOLOv4-tiny-Tea模型F1值比YOLOv3、YOLOv4、YOLOv5l模型分别提高了12.11、11.66和6.76个百分点,参数量仅为3种网络模型的13.57%、13.06%和35.05%。试验结果表明,YOLOv4-tiny-Tea模型能有效提高不同尺度下嫩芽检测的精确率,大幅度减少小尺寸或遮挡嫩芽的漏检情况,在保持轻量化计算成本的基础上获得较为明显的检测精度,能够满足农业机器人的实时检测和嵌入式开发的需求,可以为茶叶嫩芽智能采摘方法提供参考。

关键词: 茶叶, 嫩芽检测, YOLOv4-tiny, 注意力机制, 双向特征金字塔

Abstract: Precise detection of tea buds is a prerequisite for intelligent mechanical picking of tea. Aiming at the problems of poor salience and high missed detection rate of small-scale buds caused by different sizes of tea leaves and the cover of other tea leaves, this paper proposed a kind of tea buds detection model based on improved YOLOv4-tiny. In this model, a 52×52 shallow feature layer was added in the neck network to promote the attention of YOLOv4-tiny network to small target buds. A convolutional block attention module (CBAM) was introduced to suppress the background noise and improve the salience of buds, and a bidirectional feature pyramid network (BiFPN) was used to integrate characteristic information of different scales, so as to propose the YOLOv4-tiny-Tea, a high performance light weight tea buds detection model. The results of model training and performance testing on the same training set and test set show that for the YOLOv4-tiny-Tea model, the detection precision and recall rate were 97.77% and 95.23% respectively, which were 5.58% and 23.14% higher than those before modification. An ablation experiment verified the effectiveness of the modified network structure in detecting different scales of buds, and a comparison of YOLOv4-tiny-Tea model with three YOLO algorithms found that the F1 value of YOLOv4-tiny-Tea model was 12.11%, 11.66% and 6.76% higher than F1 values of YOLOv3, YOLOv4 and YOLOv5l models respectively. The number of parameters in YOLOv4-tiny-Tea model was merely 13.57%, 13.06% and 35.05% of the three network models. The experimental results demonstrate that the method proposed in this paper effectively improved the detection precision of buds under different scales, greatly reduced the missed detection rate of buds for small size or under shading, and significantly bettered the detection precision based on a lightweight computation overhead. Therefore, the method can meet the needs of agricultural robots for real-time detection and embedded development, thus providing a reference for intelligent tea buds picking.

Key words: tea, tea buds detection, YOLOv4-tiny, attention mechanism, bidirectional feature pyramid

中图分类号:

S571.1
Q126

方梦瑞, 吕军, 阮建云, 边磊, 武传宇, 姚青. 基于改进YOLOv4-tiny的茶叶嫩芽检测模型[J]. 茶叶科学, 2022, 42(4): 549-560. doi: 10.13305/j.cnki.jts.2022.04.009.

FANG Mengrui, LÜ Jun, RUAN Jianyun, BIAN Lei, WU Chuanyu, YAO Qing. Tea Buds Detection Model Using Improved YOLOv4-tiny[J]. Journal of Tea Science, 2022, 42(4): 549-560. doi: 10.13305/j.cnki.jts.2022.04.009.

参考文献

[1] 张浩, 陈勇, 汪巍, 等. 基于主动计算机视觉的茶叶采摘定位技术[J]. 农业机械学报, 2014, 45(9): 61-65.
Zhang H, Chen Y, Wang W, et al.Positioning method for tea picking using active computer vision[J]. Transactions of the Chinese Society of Agricultural Machinery, 2014, 45(9): 61-65.
[2] Chen Y T, Chen S F.Localizing plucking points of tea leaves using deep convolutional neural networks[J]. Computers and Electronics in Agriculture, 2020, 171: 105298. doi: 10.1016/j.compag.2020.105298.
[3] 张金炎, 曹成茂, 李文宝, 等. 基于多特征融合的茶叶鲜叶等级识别的方法研究[J]. 安徽农业大学学报, 2021, 48(3): 480-487.
Zhang J Y, Cao C M, Li W B, et al.Study on the method of recognition of fresh leaf grade of tea based on multi-featured fusion[J]. Journal of Anhui Agricultural University, 2021, 48(3): 480-487.
[4] Yuwana R S, Fauziah F, Heryana A, et al.Data augmentation using adversarial networks for tea diseases detection[J]. Journal Elektronika dan Telekomunikasi, 2020, 20(1): 29-35.
[5] 刘自强, 周铁军, 傅冬, 等. 基于颜色和形状的鲜茶叶图像特征提取及在茶树品种识别中的应用[J]. 江苏农业科学, 2021, 49(12): 168-172.
Liu Z Q, Zhou T J, Fu D, et al.Study on image feature extraction of fresh tea based on color and shape and its application in tea variety recognition[J]. Jiangsu Agricultural Sciences, 2021, 49(12): 168-172.
[6] 毛腾跃, 张雯娟, 帖军. 基于显著性检测和Grabcut算法的茶叶嫩芽图像分割[J]. 中南民族大学学报(自然科学版), 2021, 40(1): 80-88.
Mao T Y, Zhang W J, Tie J.Image segmentation of tea buds based on salient object detection and Grabcut[J]. Journal of South-Central Minzu University (Natural Science Edition), 2021, 40(1): 80-88.
[7] 姜苗苗, 问美倩, 周宇, 等. 基于颜色因子与图像融合的茶叶嫩芽检测方法[J]. 农业装备与车辆工程, 2020, 58(10): 44-47.
Jiang M M, Wen M Q, Zhou Y, et al.Tea bud detection method based on color factor and image fusion[J]. Agricultural Equipment & Vehicle Engineering, 2020, 58(10): 44-47.
[8] Wang T, Zhang K M, Zhang W, et al.Tea picking point detection and location based on Mask-RCNN[J]. Information Processing in Agriculture, 2021. doi: 10.1016/j.inpa.2021.12.004.
[9] Iswanto B H, Alma A .Texture histogram features for tea leaf identification using visible digital camera[J]. IOP Conference Series: Materials Science and Engineering, 2021, 1098(3): 1098-1104.
[10] 龙樟, 姜倩, 王健, 等. 茶叶嫩芽视觉识别与采摘点定位方法研究[J]. 传感器与微系统, 2022, 41(2): 39-41.
Long Z, Jiang Q, Wang J, et al.Research on method of tea flushes vision recognition and picking point localization[J]. Transducer and Microsystem Technologies, 2022, 41(2): 39-41.
[11] 吴雪梅, 张富贵, 吕敬堂. 基于图像颜色信息的茶叶嫩叶识别方法研究[J]. 茶叶科学, 2013, 33(6): 584-589.
Wu X M, Zhang F G, Lv J T.Research on recognition of tea tender leaf based on image color information[J]. Journal of Tea Science, 2013, 33(6): 584-589.
[12] 汪建. 结合颜色和区域生长的茶叶图像分割算法研究[J]. 茶叶科学, 2011, 31(1): 72-77.
Wang J.Segmentation algorithm of tea combined with the color and region growing[J]. Journal of Tea Science, 2011, 31(1): 72-77.
[13] Zhang L, Zou L, Wu C, et al.Method of famous tea sprout identification and segmentation based on improved watershed algorithm[J]. Computers and Electronics in Agriculture, 2021, 184(1): 106108. doi: 10.1016/j.compag.2021.106108.
[14] 王子钰, 赵怡巍, 刘振宇. 基于SSD算法的茶叶嫩芽检测研究[J]. 微处理机, 2020, 41(4): 42-48.
Wang Z Y, Zhao Y W, Liu Z Y.Research on tea buds detection based on SSD algorithm[J]. Microprocessors, 2020, 41(4): 42-48.
[15] 孙肖肖, 牟少敏, 许永玉, 等. 基于深度学习的复杂背景下茶叶嫩芽检测算法[J]. 河北大学学报(自然科学版), 2019, 39(2): 211-216.
Sun X X, Mu S M, Xu Y Y, et al.Detection algorithm of tea tender buds under complex background based on deep learning[J]. Journal of Hebei University (Natural Science Edition), 2019, 39(2): 211-216.
[16] Yang H, Chen L, Chen M, et al.Tender tea shoots recognition and positioning for picking robot using improved YOLO-V3 model[J]. IEEE Access, 2019: 180998-181011.
[17] Li Y T, He L Y, Jia J M, et al.In-field tea shoot detection and 3D localization using an RGB-D camera[J]. Computers and Electronics in Agriculture, 2021, 185: 106149. doi: 10.1016/j.compag.2021.106149.
[18] 吕军, 方梦瑞, 姚青, 等. 基于区域亮度自适应校正的茶叶嫩芽检测模型[J]. 农业工程学报, 2021, 37(22): 278-285.
Lyu J, Fang M R, Yao Q, et al.Detection model for tea buds based on region brightness adaptive correction[J]. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(22): 278-285.
[19] Karunasena G, Priyankara H.Tea bud leaf identification by using machine learning and image processing techniques[J]. International Journal of Scientific & Engineering Research, 2020, 11(8): 624-628.
[20] Li X, Pan J, Xie F, et al.Fast and accurate green pepper detection in complex backgrounds via an improved Yolov4-tiny model[J]. Computers and Electronics in Agriculture, 2021, 191: 106503.
[21] Jiang Z, Zhao L, Li S, et al.Real-time object detection method based on improved YOLOv4-tiny[J]. arXiv preprint, 2020, arXiv: 2011.04244. doi: 10.48550/arXiv.2011.04244.
[22] Misra D.Mish: A self regularized non-monotonic activation function[J]. arXiv preprint, 2019, arXiv: 1908.08681. doi: 10.48550/arXiv.1908.08681.
[23] Glorot X, Bordes A, Bengio Y.Deep sparse rectifier neural networks[C]//Proceedings of the fourteenth international conference on artificial intelligence and statistics. JMLR Workshop and Conference Proceedings, 2011: 315-323.
[24] Zheng Z, Wang P, Liu W, et al.Distance-IoU loss: faster and better learning for bounding box regression[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2020, 34(7): 12993-13000.
[25] Lin T Y, Dollar P, Girshick R, et al.Feature pyramid networks for object detection[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2117-2125.
[26] Woo S, Park J, Lee J Y, et al.CBAM: convolutional block attention module[C]//Proceedings of the European conference on computer vision (ECCV). 2018: 3-19.
[27] Guo C, Fan B, Zhang Q, et al.AugFPN: improving multi-scale feature learning for object detection[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020: 12595-12604.
[28] Syazwany N S, Nam J H, Lee S C.MM-BiFPN: multi-modality fusion network with Bi-FPN for MRI brain tumor segmentation[J]. IEEE Access, 2021: 160708-160720.
[29] 王金鹏, 高凯, 姜洪喆, 等. 基于改进的轻量化卷积神经网络火龙果检测方法[J]. 农业工程学报, 2020, 36(20): 218-225.
Wang J P, Gao K, Jiang H Z, et al.Method for detecting dragon fruit based on improved lightweight convolutional neural network[J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(20): 218-225.
[30] Everingham M, Van Gool L, Williams C K I, et al. The pascal visual object classes challenge[J]. International Journal of Computer Vision, 2010, 88(2): 303-338.
[31] 林森, 刘美怡, 陶志勇. 采用注意力机制与改进YOLOv5的水下珍品检测[J]. 农业工程学报, 2021, 37(18): 307-314.
Lin S, Liu M Y, Tao Z Y.Detection of underwater treasures using attention mechanism and improved YOLOv5[J]. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(18): 307-314.
[32] Redmon J, Farhadi A. YOLOv3: an incremental improvement [J]. arXiv preprint, 2018, arXiv: 1804.02767. doi.org/10.48550/arXiv.1804.02767.
[33] Bochkovskiy A, Wang C Y, Liao H Y M. Yolov4: optimal speed and accuracy of object detection[J]. arXiv preprint, 2020, arXiv: 2004.10934. doi: 10.48550/arXiv.2004.10934.
[34] Yap M H, Hachiuma R, Alavi A, et al.Deep learning in diabetic foot ulcers detection: a comprehensive evaluation[J]. Computers in Biology and Medicine, 2021, 135: 104596. doi: 10.1016/j.compbiomed.2021.104596.
[35] Selvaraju R R, Cogswell M, Das A, et al.Grad-cam: visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision. 2017: 618-626.