Lightweight fresh tea leaf recognition method based on improved YOLOv5s

doi:10.12398/j.issn.2096-7217.2025.01.001

Abstract

Abstract:

The classification and recognition of tea buds represents a crucial aspect of renowned tea production. In view of the problems of large model size， large computational complexity and inability to distinguish the picking morphology of the current tea bud recognition algorithm， this study proposes an enhanced fresh tea leaf recognition model （YOLOv5s-SPCS） based on YOLOv5s as the foundational model. Firstly， images of fresh tea leaves were collected in both laboratory and natural environments to create a dataset of fresh tea leaves. This was done through offline and online collection of images in multiple scenarios， with the resulting images divided into a training set and a test set. Secondly， the Shuffle Block module was constructed based on the ShuffleNetV2 idea for replacing the convolution module in YOLOv5s backbone network， which reduced the number of model parameters and the amount of computation while increasing the speed of feature extraction. Subsequently， the Partial Convolution structure， PConv and SimAM were incorporated into the neck network to construct the C3-PCS module， replacing the original C3 structure which further reduced the model computational redundancy and memory access， while improving the recognition accuracy with a minimal increase in the number of parameters. Finally， the SIoU bounding box loss function was employed to enhance the convergence velocity and precision of the prediction frame. In addition to accelerating the convergence of the model prediction frame regression， the use of this loss function also generates more accurately positioned prediction frames. The experimental results demonstrate that the enhanced YOLOv5s-SPCS model exhibits 14%， 14% and 16% of the YOLOv5s model in terms of the number of parameters， computational volume and weight file. The size of the model is， respectively， with an accuracy of 81.8% and a mean average precision （mAP） of 82.4% for the fresh tea image recognition， which is 2.7% more accurate than the original model. The accuracy was enhanced by 2.7 percentage points， while the mean average precision of mAP remained unaltered. Furthermore， the overall performance of the enhanced YOLOv5s-SPCS model is superior to that of the prevailing target detection models， including Faster R-CNN， SSD， YOLOv3， and YOLOv4. This study offers a valuable technical foundation for fresh tea leaves recognition classification and subsequent mobile deployment.

Key words: deep learning, YOLOv5s, lightweight, fresh tea, target detection

CLC Number:

WU Qing, WEI Runxuan, ZHOU Le, YANG Hao, LIU Wanru, XU Hongmei. Lightweight fresh tea leaf recognition method based on improved YOLOv5s[J]. Journal of Intelligent Agricultural Mechanization, 2025, 6(1): 1-14.

Figures/Tables 22

References 37

1	梅宇，张朔. 2022年中国茶叶生产与内销形势分析［J］. 中国茶叶， 2023， 45（4）： 25-30.
	MEI Yu， ZHANG Shuo. Analysis of China's tea production and domestic sales in 2022 ［J］. China Tea， 2023， 45（4）： 25-30.
2	高震宇，王安，刘勇，等. 基于卷积神经网络的鲜茶叶智能分选系统研究［J］. 农业机械学报， 2017， 48（7）： 53-58.
	GAO Zhenyu， WANG An， LIU Yong， et al. Intelligent fresh-tea-leaves sorting system research based on convolution neural network ［J］. Transactions of the Chinese Society for Agricultural Machinery， 2017， 48（7）： 53-58.
3	韩余，宋志禹，陈巧敏. 4CJ-1200F智能采茶机设计与试验［J］. 智能化农业装备学报（中英文）， 2022， 3（1）： 1-6.
	HAN Yu， SONG Zhiyu， CHEN Qiaomin. Design and experiment of 4CJ-1200F intelligent tea plucking machine ［J］. Journal of Intelligent Agricultural Mechanization， 2022， 3（1）： 1-6.
4	胡程喜，谭立新，王文胤，等. 基于改进DeepLabV3+的轻量化茶叶嫩芽采摘点识别模型［J］. 智慧农业（中英文）， 2024， 6（5）： 119-127.
	HU Chengxi， TAN Lixin， WANG Wenyin， et al. Lightweight tea shoot picking point recognition model based on improved DeepLabV3+ ［J］. Smart Agriculture， 2024， 6（5）： 119-127.
5	张兰兰，董迹芬，唐萌，等. 名优茶机采鲜叶分级技术研究［J］. 浙江大学学报（农业与生命科学版）， 2012， 38（5）： 593-598.
	ZHANG Lanlan， DONG Jifen， TANG Meng， et al. Classification technology of machine-plucking high quality tea ［J］. Journal of Zhejiang University （Agriculture and Life Sciences）， 2012， 38（5）： 593-598.
6	唐小林，李文萃，范起业. 机采茶鲜叶分类分级技术及相关设备研究进展［J］. 中国茶叶加工， 2015（2）： 5-8.
	TANG Xiaolin， LI Wencui， FAN Qiye. Research on the classification technology and the grading equipment of machine-plucking fresh leaves ［J］. China Tea Processing， 2015（2）： 5-8.
7	孙道宗，张振宇，陈俊聪，等. 一种基于深度学习的端到端生菜无损鲜重估测模型的建立［J］. 南京农业大学学报， 2024， 47（6）： 1212-1220．
	SUN Daozong， ZHANG Zhenyu， CHEN Juncong， et al ． A model for end-to-end non-destructive fresh weight estimation of lettuce based on deep learning ［J］． Journal of Nanjing Agricultural University， 2024， 47（6）： 1212-1220．
8	黄印. 基于机器学习的鲜茶叶分类研究［D］. 武汉：中南民族大学， 2021.
	HUANG Yin. Research on classification of fresh tea based on machine learning ［D］. Wuhan： South-Central Minzu University， 2021.
9	范婷婷. 基于高光谱成像技术的茶叶无损检测方法研究［D］. 天津：河北工业大学， 2021.
	FAN Tingting. Research on nondestructive detection method of tea based on hyperspectral imaging technology ［D］. Tianjin： Hebei University of Technology， 2021.
10	陈忠辉. 基于计算机视觉的茶叶嫩芽识别方法研究［D］. 贵阳：贵州大学， 2022.
	CHEN Zhonghui. Research on the method of tea sprout recognition based on computer vision ［D］. Guiyang： Guizhou University， 2022.
11	邢洁洁，谢定进，杨然兵，等. 基于YOLOv5s的农田垃圾轻量化检测方法［J］. 农业工程学报， 2022， 38（19）： 153-161.
	XING Jiejie， XIE Dingjin， YANG Ranbing， et al. Lightweight detection method for farmland waste based on YOLOv5s ［J］. Transactions of the Chinese Society of Agricultural Engineering， 2022， 38（19）： 153-161.
12	SHANG Y Y， XU X S， JIAO Y T， et al. Using lightweight deep learning algorithm for real-time detection of apple flowers in natural environments ［J］. Computers and Electronics in Agriculture， 2023， 207： 107765.
13	黄家才，唐安，陈光明，等. 基于Compact-YOLO v4的茶叶嫩芽移动端识别方法［J］. 农业机械学报， 2023， 54（3）： 282-290.
	HUANG Jiacai， TANG An， CHEN Guangming， et al. Mobile recognition solution of tea buds based on Compact-YOLO v4 algorithm ［J］. Transactions of the Chinese Society for Agricultural Machinery， 2023， 54（3）： 282-290.
14	ZHU L X， ZHANG Z H， LIN G C， et al. Detection and localization of tea bud based on improved YOLOv5s and 3D point cloud processing ［J］. Agronomy， 2023， 13（9）： 2412.
15	GUI Z Y， CHEN J N， LI Y， et al. A lightweight tea bud detection model based on Yolov5 ［J］. Computers and Electronics in Agriculture， 2023， 205： 107636.
16	王梦妮，顾寄南，王化佳，等. 基于改进YOLOv5s模型的茶叶嫩芽识别方法［J］. 农业工程学报， 2023， 39（12）： 150-157.
	WANG Mengni， GU Jinan， WANG Huajia， et al. Method for identifying tea buds based on improved YOLOv5s model ［J］. Transactions of the Chinese Society of Agricultural Engineering， 2023， 39（12）： 150-157.
17	白强，高荣华，赵春江，等. 基于改进YOLOV5s网络的奶牛多尺度行为识别方法［J］. 农业工程学报， 2022， 38（12）： 163-172.
	BAI Qiang， GAO Ronghua， ZHAO Chunjiang， et al. Multi-scale behavior recognition method for dairy cows based on improved YOLOV5s network ［J］. Transactions of the Chinese Society of Agricultural Engineering， 2022， 38（12）： 163-172.
18	WOLPERT D H， MACREADY W G. No free lunch theorems for optimization ［J］. IEEE Transactions on Evolutionary Computation， 1997， 1（1）： 67-82.
19	LIN T Y， DOLLÁR P， GIRSHICK R， et al. Feature pyramid networks for object detection ［C］// 2017 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway， New Jersey： IEEE， 2017： 936-944.
20	LIU S， QI L， QIN H F， et al. Path aggregation network for instance segmentation ［C］// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway， New Jersey： IEEE， 2018： 8759-8768.
21	MA N N， ZHANG X Y， ZHENG H T， et al. ShufflenetV2： Practical guidelines for efficient CNN architecture design ［C］//Proceedings of the European Conference on Computer Vision （ECCV）. 2018： 116-131.
22	ZHANG X Y， ZHOU X Y， LIN M X， et al. Shufflenet： An extremely efficient convolutional neural network for mobile devices ［C］//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018： 6848-6856.
23	CHEN J R， KAO S H， HE H， et al. Run， Don't walk： Chasing higher FLOPS for faster neural networks ［C］// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023： 12021-12031.
24	YANG L X， ZHANG R Y， LI L D， et al. SimAM： A simple， parameter-free attention module for convolutional neural networks ［C］// International Conference on Machine Learning. PMLR， 2021： 11863-11874.
25	GEVORGYAN Z. SIoU loss： More powerful learning for bounding box regression ［J］. arXiv preprint arXiv：， 2022.
26	HOWARD A， SANDLER M， CHEN B， et al. Searching for MobileNetV3 ［C］// 2019 IEEE/CVF International Conference on Computer Vision （ICCV）. Piscataway， New Jersey： IEEE， 2019： 1314-1324.
27	HAN K， WANG Y H， TIAN Q， et al. GhostNet： More features from cheap operations ［C］// 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway， New Jersey： IEEE， 2020： 1577-1586.
28	REZATOFIGHI H， TSOI N， GWAK J， et al. Generalized intersection over union： A metric and a loss for bounding box regression ［C］// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway， New Jersey： IEEE， 2019： 658-666.
29	ZHENG Z H， WANG P， REN D W， et al. Enhancing geometric factors in model learning and inference for object detection and instance segmentation ［J］. IEEE Transactions on Cybernetics， 2022， 52（8）： 8574-8586.
30	ZHANG Y F， REN W Q， ZHANG Z， et al. Focal and efficient IoU loss for accurate bounding box regression ［J］. Neurocomputing， 2022， 506： 146-157.
31	REN S Q， HE K M， GIRSHICK R， et al. Faster R-CNN： Towards real-time object detection with region proposal networks ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2016， 39（6）： 1137-1149.
32	周华茂，王婧，殷华，等. 基于改进 Mask R-CNN 模型的秀珍菇表型参数自动测量方法［J］. 智慧农业（中英文）， 2023， 5（4）： 117-126.
	ZHOU Huamao， WANG Jing， YIN Hua， et al. Phenotype analysis of Pleurotus geesteranus based on improved Mask R-CNN ［J］. Smart Agriculture， 2023， 5（4）： 117-126.
33	郭小燕，于帅卿．一种轻量级 YOLOv5S 农作物虫害目标检测模型［J］．南京农业大学学报，2024，47（5）： 1009-1018．
	GUO Xiaoyan， YU Shuaiqing ． A lightweight YOLOv5S crop pest target detection model［J］． Journal of Nanjing Agricultural University， 2024， 47（5）： 1009-1018．
34	REDMON J， FARHADI A. Yolov3： An incremental improvement ［J］. arXiv preprint arXiv：， 2018.
35	BOCHKOVSKIY A， WANG C Y， LIAO H Y M. Yolov4： Optimal speed and accuracy of object detection ［J］. arXiv preprint arXiv：， 2020.
36	WANG C Y， BOCHKOVSKIY A， LIAO H Y M. YOLOv7： Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors ［C］// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway， New Jersey： IEEE， 2023： 7464-7475.
37	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.

模型	深度	宽度	参数量 /×10⁶ M	浮点运算量 FLOPs/G
YOLOv5n	0.33	0.25	1.76	4.1
YOLOv5s	0.33	0.50	7.02	15.8
YOLOv5m	0.67	0.75	20.86	47.9
YOLOv5l	1.00	1.00	46.15	108.3
YOLOv5x	1.33	1.25	86.23	204.7

模型	深度	宽度	参数量 /×10⁶ M	浮点运算量 FLOPs/G
YOLOv5n	0.33	0.25	1.76	4.1
YOLOv5s	0.33	0.50	7.02	15.8
YOLOv5m	0.67	0.75	20.86	47.9
YOLOv5l	1.00	1.00	46.15	108.3
YOLOv5x	1.33	1.25	86.23	204.7

Shuffle Block	PConv	SimAM	SIoU	精确度 P/%	召回率 R/%	平均精度均值 mAP/%	参数量 Params/×10⁶ M	浮点运算量 FLOPs/G	权重文件大小 /MB
×	×	×	×	79.1	72.1	82.4	7.02	15.8	13.67
√	×	×	×	78.2	69.1	80.4	0.85	1.8	1.85
×	√	×	×	82.3	72.3	82.1	7.20	16	14.03
×	×	√	×	82.1	72.7	82.5	7.29	17.2	14.20
√	√	×	×	79.5	68.2	81.4	0.89	1.9	1.95
×	√	√	×	80.9	73.1	82.6	7.48	17.4	14.56
√	×	√	×	78.1	70.3	80.9	0.91	2.2	1.99
√	√	√	×	81.8	71.9	81.2	0.96	2.2	2.09
√	√	√	√	81.8	72.5	82.4	0.96	2.2	2.09

Shuffle Block	PConv	SimAM	SIoU	精确度 P/%	召回率 R/%	平均精度均值 mAP/%	参数量 Params/×10⁶ M	浮点运算量 FLOPs/G	权重文件大小 /MB
×	×	×	×	79.1	72.1	82.4	7.02	15.8	13.67
√	×	×	×	78.2	69.1	80.4	0.85	1.8	1.85
×	√	×	×	82.3	72.3	82.1	7.20	16	14.03
×	×	√	×	82.1	72.7	82.5	7.29	17.2	14.20
√	√	×	×	79.5	68.2	81.4	0.89	1.9	1.95
×	√	√	×	80.9	73.1	82.6	7.48	17.4	14.56
√	×	√	×	78.1	70.3	80.9	0.91	2.2	1.99
√	√	√	×	81.8	71.9	81.2	0.96	2.2	2.09
√	√	√	√	81.8	72.5	82.4	0.96	2.2	2.09

类别	精确率 P/%	召回率 R/%	平均精度均值 mAP/%
全类别	81.8	72.5	82.4
一芽一叶	79.5	75.5	85.3
一芽二叶	86.5	79.1	89.4
单芽	79.9	61.2	72.1