Development of a yellow sticky trap pest identification platform for open-field vegetables

doi:10.16178/j.issn.0528-9017.20250569

Abstract

Abstract:

Pest infestation in open fields has long been a critical factor constraining the high-quality development of China's vegetable industry. Timely and effective pest early warning is of great significance for ensuring vegetable production safety. This study established a yellow sticky trap dataset featuring 5 common vegetable pests, including Bemisia tabaci, Bactrocera tau, Hymenia recurvalis, Phyllotreta striolata, and Cicadellidae by collecting field data using a “field+darkbox” imaging approach. Based on this dataset, YOLO-series models were trained and compared for pest detection performance. The results showed that YOLOv11s performed the best, achieving a mean average precision (mAP50) of 0.973. The trained model was deployed as a WeChat Mini Program called “PhenoAI”, providing the public with pest identification services and offering a more practical technical solution for field pest control.

Key words: open-field vegetable, pest identification, YOLO-series model, online detection

CLC Number:

CHU Hangjian, LOU Weidong, GU Qing, LIU Qing, WU Hongyang, HUANG Xiaohua, ZHANG Xiaobin. Development of a yellow sticky trap pest identification platform for open-field vegetables[J]. Journal of Zhejiang Agricultural Sciences, 2026, 67(1): 136-142.

Figures/Tables 7

References 19

[1]	吴霜, 张谊模, 郭萧, 等. 不同波长LED诱虫灯对菜地昆虫的诱集效果[J]. 应用昆虫学报, 2021, 58(1): 172-180.
	WU S, ZHANG Y M, GUO X, et al. Effectiveness of different wavelength LED insect lamp traps in vegetable fields[J]. Chinese Journal of Applied Entomology, 2021, 58(1): 172-180.
[2]	黄洁珧, 黄伟康, 秦双, 等. 7种生物源杀虫剂对烟粉虱杀虫活性评价与解毒酶响应机制研究[J]. 中国瓜菜, 2025, 38(4): 166-173.
	HUANG J Y, HUANG W K, QIN S, et al. Evaluation of insecticidal activity and detoxification enzyme response mechanism of seven biopesticides against Bemisia tabaci(Gennadius)[J]. China Cucurbits and Vegetables, 2025, 38(4): 166-173.
[3]	田晓丽, 许冬, 刘毅, 等. 南亚果实蝇谷胱甘肽S-转移酶基因的鉴定及组织表达[J]. 西北农林科技大学学报(自然科学版), 2025, 53(10): 122-131.
	TIAN X L, XU D, LIU Y, et al. Identification and tissue expression of glutathione S-transferase gene in Bactrocera tau[J]. Journal of Northwest A & F University (Natural Science Edition), 2025, 53(10): 122-131.
[4]	徐嘉政, 占绣萍, 田椿燕, 等. 黄曲条跳甲的防治药剂筛选及其对药剂的季节敏感性研究[J]. 中国植保导刊, 2023, 43(6): 5-9.
	XU J Z, ZHAN X P, TIAN C Y, et al. Screening of pesticides for Phyllotreta striolata control and the seasonal sensitivity of this insect to pesticides[J]. China Plant Protection, 2023, 43(6): 5-9.
[5]	田鑫钰. 稻田主要害虫在线智能识别系统的应用研究[D]. 泰安: 山东农业大学, 2023.
	TIAN X Y. Application research on an online intelligent recognition system for major pests in rice fields[D]. Tai'an: Shandong Agricultural University, 2023.
[6]	李凌, 朱健文, 杨欢. 基于YOLOv8改进的农业复杂场景害虫识别算法[J/OL]. 软件导刊, 2025: 1-9. ( 2025-05-22). https://kns.cnki.net/kcms/detail/detail.aspx?dbcode=CJFD&dbname=CJFD&filename=RJDK20250908008.
	LI L, ZHU J W, YANG H. Pest identification algorithm in complex agricultural scenes based on improved YOLOv8[J/OL]. Software Guide, 2025: 1-9. ( 2025-05-22). https://kns.cnki.net/kcms/detail/detail.aspx?dbcode=CJFD&dbname=CJFD&filename=RJDK20250908008.
[7]	王海亮. 面向农田实际场景害虫识别的专用数据集构建及智能检测方法研究[D]. 杭州: 杭州电子科技大学, 2024.
	WANG H L. Construction of a dedicated dataset and intelligent detection methods for pest recognition in agricultural fields[D]. Hangzhou: Hangzhou Dianzi University, 2024.
[8]	李雨晴, 陈燕红, 李永可, 等. 作物害虫图像智能识别方法[J]. 新疆农业科学, 2023, 60(12): 2973-2981.
	LI Y Q, CHEN Y H, LI Y K, et al. Image intelligent recognition method of crop pests[J]. Xinjiang Agricultural Sciences, 2023, 60(12): 2973-2981.
[9]	潘春华. 基于机器学习的南方蔬菜重大害虫识别预警研究[D]. 广州: 华南农业大学, 2019.
	PAN C H. Recognition and early warning for major vegetable pests in South China based on machine learning[D]. Guangzhou: South China Agricultural University, 2019.
[10]	刘珍, 王菲菲, 李勇华, 等. 智能识别虫情测报灯在蔬菜害虫测报上的应用效果[J]. 中国植保导刊, 2022, 42(8): 37-41.
	LIU Z, WANG F F, LI Y H, et al. Application effect of intelligent pest identification forecasting lamp in vegetable pest forecasting[J]. China Plant Protection, 2022, 42(8): 37-41.
[11]	LI W Y, WANG D J, LI M, et al. Field detection of tiny pests from sticky trap images using deep learning in agricultural greenhouse[J]. Computers and Electronics in Agriculture, 2021, 183: 106048.
[12]	卜俊怡. 基于机器视觉的温室诱虫板害虫图像自动识别系统设计[D]. 南京: 南京农业大学, 2022.
	BU J Y. Design on a machine vision-based automatic identification system for pest trapped on yellow sticky papers in greenhouse[D]. Nanjing: Nanjing Agricultural University, 2022.
[13]	KALFAS I, DE KETELAERE B, BUNKENS K, et al. Towards automatic insect monitoring on witloof chicory fields using sticky plate image analysis[J]. Ecological Informatics, 2023, 75: 102037.
[14]	周国涛, 李明珠, 张建涛, 等. 智能虫情测报灯装置优化与应用效果[J]. 中国植保导刊, 2023, 43(10): 48-53.
	ZHOU G T, LI M Z, ZHANG J T, et al. Development and application of intelligent pest forecasting light[J]. China Plant Protection, 2023, 43(10): 48-53.
[15]	TERVEN J, CÓRDOVA-ESPARZA D M, ROMERO-GONZÁLEZ J A. A comprehensive review of YOLO architectures in computer vision: from YOLOv1 to YOLOv8 and YOLO-NAS[J]. Machine Learning and Knowledge Extraction, 2023, 5(4): 1680-1716.
[16]	SAPKOTA R, CALERO M F, QURESHI R, et al. YOLO advances to its genesis: a decadal and comprehensive review of the You Only Look Once (YOLO) series[J]. Artificial Intelligence Review, 2025, 58: 274.
[17]	DIWAN T, ANIRUDH G, TEMBHURNE J V. Object detection using YOLO: challenges, architectural successors, datasets and applications[J]. Multimedia Tools and Applications, 2023, 82(6): 9243-9275.
[18]	JIANG P Y, ERGU D, LIU F Y, et al. A review of yolo algorithm developments[J]. Procedia Computer Science, 2022, 199: 1066-1073.
[19]	Ultralytics. Ultralytics YOLO docs: YOLO-a brief history[EB/OL].[2025-07-11]. https://docs.ultralytics.com/zh/#yolo-a-brief-history.

害虫名称	训练集目标框数量	验证集目标框数量
烟粉虱	736	178
南亚果实蝇	1 141	307
甜菜白带野螟	796	210
黄曲条跳甲	963	189
叶蝉	862	228

害虫名称	训练集目标框数量	验证集目标框数量
烟粉虱	736	178
南亚果实蝇	1 141	307
甜菜白带野螟	796	210
黄曲条跳甲	963	189
叶蝉	862	228

YOLO模型	精确度	召回率	mAP50	mAP50-95
v5n	0.937	0.935	0.965	0.726
v5s	0.934	0.926	0.963	0.756
v8n	0.919	0.958	0.967	0.738
v8s	0.932	0.940	0.968	0.763
v11n	0.948	0.939	0.975	0.753
v11s	0.932	0.959	0.973	0.779

YOLO模型	精确度	召回率	mAP50	mAP50-95
v5n	0.937	0.935	0.965	0.726
v5s	0.934	0.926	0.963	0.756
v8n	0.919	0.958	0.967	0.738
v8s	0.932	0.940	0.968	0.763
v11n	0.948	0.939	0.975	0.753
v11s	0.932	0.959	0.973	0.779

类别	标注信息	精确度/%	召回率/%	mAP50/%	mAP50- 95/%
烟粉虱	178	0.889	0.959	0.952	0.745
南亚果实蝇	307	0.911	0.944	0.957	0.767
甜菜白带野螟	210	0.976	0.928	0.993	0.861
黄曲条跳甲	189	0.963	0.985	0.995	0.861
叶蝉	228	0.921	0.995	0.967	0.663
总计	1 112	0.932	0.959	0.973	0.779