Design and experiment of tea bud recognition method and light source of recognition system based on YOLOv5-SPD

doi:10.12398/j.issn.2096-7217.2024.03.004

Abstract

Abstract:

The realization of automatic identification and positioning of tea buds is the basis for the development of high-quality tea intelligent picking equipment. Aiming at the problems of tiny tea buds and the picking environment which is greatly affected by light， this research proposes a tea bud recognition method based on a deep learning network model to carry out the light source design of the recognition system， which can provide technical support for the realization of all-weather and high-efficiency intelligent tea bud picking equipment. First， an aluminum alloy frame was constructed to provide a closed and shaded dark environment； then， three combinations of heights and three combinations of light intensities were created by adjusting the height of the crossbar and the brightness of the light source； finally， the image datasets of tea buds in different combinations were collected， and recognition tests were carried out on one bud and one leaf and one bud and two leaves by using the improved YOLOv5 model. The experimental results show that the overall accuracy of YOLOv5s is 77.13%， and the overall average precision mean is 86.14%， the overall accuracy of the improved recognition model YOLOv5s-SPD is 80.30%， and the overall average precision mean is 87.3%， and the average detection time of a single image is 5.7 ms， which meets the requirement of real-time detection， and is better than the original YOLOv5s with an overall accuracy improvement of 3.17% and an overall average precision mean of 1.16%， which effectively improves the recognition performance of tea buds. Under the condition of height 90 cm and luminance L7（0.164~0.328 $μ m o l / m 2$ ）， the detection accuracy， recall and AP average of one bud and two leaves were 86.70%， 92.45% and 95.00%， respectively. The method proposed in this paper can effectively and quickly detect tea buds， and the light source design scheme supports the development of all-weather high-quality tea intelligent picking equipment.

Key words: tea buds, deep learning, lighting systems, intelligent picking, target detection, robotics

CLC Number:

WANG Yuanhong, YANG Zhiming, WANG Qi, LU Jinzhu, GAO Junfeng. Design and experiment of tea bud recognition method and light source of recognition system based on YOLOv5-SPD[J]. Journal of Intelligent Agricultural Mechanization, 2024, 5(3): 33-43.

Figures/Tables 12

References 25

1	AHAMMED G J， LI X. Hormonal regulation of health-promoting compounds in tea （Camellia sinensis L.）［J］. Plant Physiol Biochem， 2022， 185： 390-400.
2	DINIZ P H G D， PISTONESI M F， ALVAREZ M B， et al. Simplified tea classification based on a reduced chemical composition profile via successive projections algorithm linear discriminant analysis （SPA-LDA）［J］. Journal of Food Composition and Analysis， 2015， 39： 103-110.
3	刘勤，胡良龙，郑砚砚，等. 中国茶叶及蔬菜智能化收获技术发展现状（英文）［J］. 智能化农业装备学报（中英文）， 2021， 2（1）： 20-27.
	LIU Qin， HU Lianglong， ZHENG Yanyan， et al. Current status of the development of intelligent harvesting technology for tea and vegetables in China ［J］. Journal of Intelligent Agricultural Mechanization（In Chinese and English）， 2021， 2（1）： 20-27.
4	SHUAI L， MU J， JIANG X， et al. An improved YOLOv5-based method for multi-species tea shoot detection and picking point location in complex backgrounds ［J］. Biosystems Engineering， 2023， 231： 117-132.
5	GUI Z， CHEN J， LI Y， et al. A lightweight tea bud detection model based on Yolov5 ［J］. Computers and Electronics in Agriculture， 2023， 205： 107636.
6	熊俊涛，林睿，刘振，等. 夜间自然环境下荔枝采摘机器人识别技术［J］. 农业机械学报， 2017， 48（11）： 28-34.
	XIONG Juntao， LIN Rui， LIU Zhen， et al. Recognition technology of lychee picking robot in natural environment at night ［J］. Journal of Agricultural Machinery， 2017， 48（11）： 28-34.
7	肖张娜，罗陆锋，陈明猷，等. 基于改进YOLO-v4的果园环境下葡萄检测［J］. 智能化农业装备学报（中英文）， 2023， 4（2）： 35-43.
	XIAO Zhangna， LUO Lufeng， CHEN Mingyou， et al. Grape detection in orchard environment based on improved YOLO-v4 ［J］. Journal of Intelligent Agricultural Equipment， 2023， 4（2）： 35-43.
8	李搴曦，孙晓明，江晗慧，等. 基于YOLOv4-tiny的设施番茄智能喷药无人车设计与试验［J］. 智能化农业装备学报（中英文）， 2023， 4（2）： 44-52.
	LI Qianxi， SUN Xiaoming， JIANG Hanhui， et al. Design and test of an unmanned vehicle for intelligent spraying of tomato based on YOLOv4-tiny ［J］. Journal of Intelligent Agricultural Equipment， 2023， 4（2）： 44-52.
9	LIU X， ZHAO D， JIA W， et al. A method of segmenting apples at night based on color and position information ［J］. Computers and Electronics in Agriculture， 2016， 122： 118-123.
10	XIANG R. Image segmentation for whole tomato plant recognition at night ［J］. Computers and Electronics in Agriculture， 2018， 154： 434-442.
11	FONT D， PALLEJÀ T， TRESANCHEZ M， et al. Counting red grapes in vineyards by detecting specular spherical reflection peaks in RGB images obtained at night with artificial illumination ［J］. Computers and Electronics in Agriculture， 2014， 108： 105-111.
12	PAYNE A， WALSH K， SUBEDI P， et al. Estimating mango crop yield using image analysis using fruit at ‘stone hardening’ stage and night time imaging ［J］. Computers and Electronics in Agriculture， 2014， 100： 160-167.
13	LIANG C， XIONG J， ZHENG Z， et al. A visual detection method for nighttime litchi fruits and fruiting stems ［J］. Computers and Electronics in Agriculture， 2020， 169： 105192.
14	LADDI A， SHARMA S， KUMAR A， et al. Classification of tea grains based upon image texture feature analysis under different illumination conditions ［J］. Journal of Food Engineering， 2013， 115（2）： 226-231.
15	LI J， LU Y， LU R. Detection of early decay in navel oranges by structured-illumination reflectance imaging combined with image enhancement and segmentation ［J］. Postharvest Biology and Technology， 2023， 196： 112162.
16	LI G， YANG Y， QU X， et al. A deep learning based image enhancement approach for autonomous driving at night ［J］. Knowledge-Based Systems， 2021， 213.
17	BOCHKOVSKIY A， WANG C Y， LIAO H Y M. Yolov4： Optimal speed and accuracy of object detection ［J］. arxiv preprint arxiv：， 2020.
18	REDMON J， FARHADI A. Yolov3： An incremental improvement ［J］. arXiv preprint arXiv：， 2018.
19	QI X， DONG J， LAN Y， et al. Method for identifying litchi picking position based on YOLOv5 and PSPNet ［J］. Remote Sensing， 2022， 14（9）： 2004.
20	LIN T Y， DOLLÁR 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.
21	LIU S， QI L， QIN H， et al. Path aggregation network for instance segmentation ［C］// Proceedings of the IEEE conference on computer vision and pattern recognition， 2018： 8759-8768.
22	WANG Z， WANG C， CHEN Y， et al. Target detection algorithm based on super-resolution color remote sensing image reconstruction ［J］. Journal of Measurements in Engineering， 2024， 12（1）： 83-98.
23	HOU Q， ZHOU D， FENG J. Coordinate attention for efficient mobile network design ［C］// Proceedings of the IEEE/CVF conference on computer vision and pattern recognition， 2021： 13713-13722.
24	CHEN J， WU J， WANG Z， et al. Detecting ripe fruits under natural occlusion and illumination conditions ［J］. Computers and Electronics in Agriculture， 2021， 190： 106450.
25	MIRHAJI H， SOLEYMANI M， ASAKEREH A， et al. Fruit detection and load estimation of an orange orchard using the YOLO models through simple approaches in different imaging and illumination conditions ［J］. Computers and Electronics in Agriculture， 2021， 191： 106533.

硬件	配置	环境	版本
System	Windows10	Python	3.10.11
CPU	Intel(R) i5-13400F	Pytorch	1.13.1
GPU	NVIDIA RTX 3060Ti	Pycharm	2023.1
RAM	32.0 GB	CUDA	11.7
Hard-disk	1 TB	CUDNN	8.4.1.5

硬件	配置	环境	版本
System	Windows10	Python	3.10.11
CPU	Intel(R) i5-13400F	Pytorch	1.13.1
GPU	NVIDIA RTX 3060Ti	Pycharm	2023.1
RAM	32.0 GB	CUDA	11.7
Hard-disk	1 TB	CUDNN	8.4.1.5

高度/cm	光强	类型	准确率/%	合并准确率/%	召回率/%	合并召回率/%	平均精度均值/%	合并平均精度均值/%
80	L6	1	76.50	70.90	82.50	73.05	84.40	78.10
	L6	2	65.30	70.90	63.60	73.05	71.80	78.10
	L7	1	84.70	83.60	90.90	88.65	90.20	87.60
	L7	2	82.50	83.60	86.40	88.65	85.00	87.60
	L8	1	81.50	77.95	89.50	79.95	92.30	82.45
	L8	2	74.40	77.95	70.40	79.95	72.60	82.45
90	L6	1	83.70	70.60	86.40	83.80	91.10	83.95
	L6	2	57.50	70.60	81.20	83.80	76.80	83.95
	L7	1	92.90	86.70	97.00	92.45	96.70	95.00
	L7	2	80.50	86.70	87.90	92.45	93.30	95.00
	L8	1	86.30	82.95	93.20	83.30	87.80	85.70
	L8	2	79.60	82.95	83.40	83.30	83.60	85.70
100	L6	1	85.90	42.95	89.40	44.70	93.60	46.80
	L6	2	0	42.95	0	44.70	0	46.80
	L7	1	89.40	82.80	93.40	90.45	95.00	93.10
	L7	2	76.20	82.80	87.50	90.45	91.20	93.10
	L8	1 2	87.50 80.70	84.10	86.80 86.40	86.60	93.80 84.90	89.35

高度/cm	光强	类型	准确率/%	合并准确率/%	召回率/%	合并召回率/%	平均精度均值/%	合并平均精度均值/%
80	L6	1	76.50	70.90	82.50	73.05	84.40	78.10
	L6	2	65.30	70.90	63.60	73.05	71.80	78.10
	L7	1	84.70	83.60	90.90	88.65	90.20	87.60
	L7	2	82.50	83.60	86.40	88.65	85.00	87.60
	L8	1	81.50	77.95	89.50	79.95	92.30	82.45
	L8	2	74.40	77.95	70.40	79.95	72.60	82.45
90	L6	1	83.70	70.60	86.40	83.80	91.10	83.95
	L6	2	57.50	70.60	81.20	83.80	76.80	83.95
	L7	1	92.90	86.70	97.00	92.45	96.70	95.00
	L7	2	80.50	86.70	87.90	92.45	93.30	95.00
	L8	1	86.30	82.95	93.20	83.30	87.80	85.70
	L8	2	79.60	82.95	83.40	83.30	83.60	85.70
100	L6	1	85.90	42.95	89.40	44.70	93.60	46.80
	L6	2	0	42.95	0	44.70	0	46.80
	L7	1	89.40	82.80	93.40	90.45	95.00	93.10
	L7	2	76.20	82.80	87.50	90.45	91.20	93.10
	L8	1 2	87.50 80.70	84.10	86.80 86.40	86.60	93.80 84.90	89.35

高度/cm	光强	类型	准确率/%	合并准确率/%	召回率/%	合并召回率/%	平均精度均值/%	合并平均精度均值/%
80	L6	1	76.10	71.35	84.50	80.75	88.90	83.75
	L6	2	66.60	71.35	77.00	80.75	78.60	83.75
	L7	1	83.30	80.85	90.60	92.75	93.30	92.35
	L7	2	78.40	80.85	94.90	92.75	91.40	92.35
	L8	1	78.60	76.50	92.40	79.55	91.70	82.75
	L8	2	74.40	76.50	66.70	79.55	73.80	82.75
90	L6	1	73.90	66.15	89.40	85.30	88.80	84.35
	L6	2	58.40	66.15	81.20	85.30	79.90	84.35
	L7	1	90.60	85.95	93.60	93.70	95.00	92.60
	L7	2	81.30	85.95	93.80	93.70	90.20	92.60
	L8	1	84.80	78.20	84.20	83.15	88.70	88.30
	L8	2	71.60	78.20	82.10	83.15	87.90	88.30
100	L6	1	79.70	39.85	98.20	49.10	93.10	46.55
	L6	2	0	39.85	0	49.10	0	46.55
	L7	1	91.30	80.80	95.80	90.30	96.20	90.75
	L7	2	70.30	80.80	84.80	90.30	85.30	90.75
	L8	1 2	87.00 74.80	80.90	88.80 86.20	87.50	94.50 89.30	91.90