基于YOLOX—Nano的稻田空心莲子草目标检测方法研究

doi:10.13733/j.jcam.issn.2095-5553.2025.07.025

摘要/Abstract

摘要： 稻田空心莲子草为稻田外来入侵恶性杂草，在秧苗封行前与水稻幼苗共同竞争水、肥、光和生长空间等资源，进而影响水稻产量。目前主要采用化学喷施除草剂对草害进行防治，如果在稻田无差别地喷施化学除草剂，不仅污染环境，还对稻田秧苗产生一定的药害。随着深度学习的快速发展，根据稻田杂草的区域位置进行除草剂精准喷施已成为可能。YOLO系列具有高效推理能力及快速迭代的特点，系列版本体积小，方便嵌入手机或其他终端产品。以秧苗未封行前稻田中的恶性杂草空心莲子草为研究对象，针对YOLOX一阶段目标检测模型系列，YOLOX—Nano虽然在YOLOX系列中性能不是最优，但是其在YOLOX几个版本中体积最小，方便嵌入手机或其他终端产品，最终选择YOLOX系列中的YOLOX—Nano模型。同时构建YOLOv3、YOLOv4—tiny、YOLOv5—s、SSD和YOLOX—Nano等空心莲子草目标检测模型，并进行性能实验对比。结果表明，基于YOLOX—Nano的空心莲子草目标检测模型中召回率Recall、mAP、F1值都高于YOLOv3、YOLOv4—tiny、SSD和YOLOv5—s模型，分别达到97.14%、96.72%和93%。针对秧苗和空心莲子草之间不同程度的轻微遮挡和部分严重遮挡图像，基于YOLOX—Nano的空心莲子草目标检测模型检测效果好于YOLOv3、YOLOv4—tiny、SSD和YOLOv5—s。

关键词: 空心莲子草, 稻田, 目标检测, 除草, YOLOX—Nano

Abstract: Alternanthera philoxeroides is an alien invasive malignant weed in rice fields, which competes with rice seedlings for water, fertilizer, light and growth space before seedling closure, thus seriously affecting rice yield. At present, chemical herbicides are mainly used to control weeds. If indiscriminate spraying of chemical herbicides is used in rice fields, it will not only polluting the environment but also causing certain phytotoxicity to rice seedlings. With the rapid development of artificial intelligence and deep learning, it is possible to precisely spray herbicides according to the regional location of weeds in rice fields. Due to the efficient reasoning ability and fast iteration of the YOLO series with one-stage target detection models, the series version has become smaller in size and can be easily embedded in mobile phones or other terminal products. This article focuses on the malignant weed hollow lotus seed grass in paddy fields before the seedlings are sealed off, and the YOLOX series are established. The performance of YOLOX Nano in the YOLOX series is not optimal. Due to its smallest size among several versions of YOLOX, which could be easily embedded into mobile phones or other terminal products, the YOLOX Nano model in the YOLOX series was ultimately chosen. The YOLOv3, YOLOv4—tiny, YOLOv5—s, SSD, and YOLOX—Nano target detection networks were constructed and their performance were compared. The experimental results showed that the Recall, mAP and F1 values of the YOLOX—Nano based target detection model of Alternanthera philoxeroides were higher than those of the YOLOv3, YOLOv4—tiny, SSD, and YOLOv5—s, reaching 97.14%, 96.72% and 93%, respectively. Aiming at the slight occlusion and partially severe occlusion image between seedlings and Alternanthera philoxeroides, the detection effect of Alternanthera philoxeroides target detection model based on YOLOX—Nano is better than that of YOLOv3, YOLOv4—tiny, SSD, and YOLOv5—s.

Key words: Alternanthera philoxeroides, paddy field, object detection, weed, YOLOX—Nano

中图分类号:

梁松, 李华锋, 邓向武, 谢新雪, 李岳鑫, 刘星晨. 基于YOLOX—Nano的稻田空心莲子草目标检测方法研究[J]. 中国农机化学报, 2025, 46(7): 173-180.

Liang Song, Li Huafeng, Deng Xiangwu, Xie Xinxue, Li Yuexin, Liu Xingchen. Research on target detection method of Alternanthera philoxeroides in rice seedling stage based on YOLOX—Nano[J]. Journal of Chinese Agricultural Mechanization, 2025, 46(7): 173-180.

参考文献

［1］　叶照春，陈仕红，冉海燕，等. 空心莲子草对草甘膦的敏感性测定［J］. 农药，2020，59（12）：925-927.〖JP2〗Ye Zhaochun, Chen Shihong, Ran Haiyan, et al. The sensitivity of Alternanthera philoxeroides to Glyphosate ［J］. Agrochemicals, 2020, 59（12）: 925-927.〖JP〗
［2］　吕沐华, 丁珠玉. 基于机器视觉的果园喷药除草机器人视觉系统设计［J］. 中国农机化学报, 2021, 42（5）: 42-48.
Lü Muhua, Ding Zhuyu. Design of visual systems for orchard spraying weeding robot based on machine vision ［J］. Journal of Chinese Agricultural Mechanization, 2021, 42（5）: 42-48.
［3］　李东升, 胡文泽, 兰玉彬, 等. 深度学习在杂草识别领域的研究现状与展望［J］. 中国农机化学报, 2022, 43（9）：137-144.
Li Dongsheng, Hu Wenze, Lan Yubin, et al. Research status and prospect of deep learning in weed recognition ［J］.Journal of Chinese Agricultural Mechanization, 2022, 43（9）: 137-144.
［4］　付豪, 赵学观, 翟长远, 等. 基于深度学习的杂草识别方法研究进展［J］. 中国农机化学报, 2023, 44 （5）: 198-207.
Fu Hao, Zhao Xueguan, Zhai Changyuan, et al. Research progress on weed recognition method based on deep learning technology ［J］. Journal of Chinese Agricultural Mechanization, 2023, 44 （5）: 198-207.
［5］　邢钦淞, 丁素明, 薛新宇, 等. 智能田间除草机器人发展现状研究［J］. 中国农机化学报, 2022, 43（8）: 173-181.
Xing Qinsong, Ding Suming, Xue Xinyu, et al. Research on the development starts of intelligent field weeding robot ［J］. Journal of Chinese Agricultural Mechanization, 2022,43（8）: 173-181.
［6］　姚思雨，王磊，张宏文. 基于CNN的棉田杂草识别方法［J］. 石河子大学学报（自然科学版）, 2023, 41（1）:21-26.
Yao Siyu, Wang Lei, Zhang Hongwen. Weed identificationmethod in cotton field based on CNN ［J］. Journal of Shihezi University （Natural Science）, 2023, 41（1）: 21-26.
［7］　胡炼, 刘海龙, 何杰, 等. 智能除草机器人研究现状与展望［J］. 华南农业大学学报, 2023, 44（1）: 34-42.
Hu Lian, Liu Hailong, He Jie, et al. Research progress and prospect of intelligent weeding robot ［J］. Journal of South China Agricultural University, 2023,44（1）: 34-42.
［8］　Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 580-587.
［9］　李晓光, 付陈平, 李晓莉, 等. 面向多尺度目标检测的改进 Faster R—CNN算法［J］. 计算机辅助设计与图形学学报, 2019, 31（7）: 1095-1101.
Li Xiaoguang, Fu Chenping, Li Xiaoli, et al. Improved Faster R—CNN for multi-scale object detection ［J］. Journal of Computer-Aided Design & Computer Graphics, 2019, 31（7）: 1095-1101.
［10］　Uijlings J R R, Sande K E A, Gevers T, et al. Selective search for object recognition ［J］. International Journal of Computer Vision, 2013, 104（2）: 154-171.
［11］　Cao C, Wang B, Zhang W, et al. An improved Faster R—CNN for small object detection ［J］. IEEE Access, 2019, 7: 106838-106846.
［12］　黄书琴，黄福乐，罗柳茗，等．基于Faster R—CNN的蔗田杂草检测算法研究［J］. 中国农机化学报，2024，45（6）:208-215.
Huang Shuqin, Huang Fule, Luo Liuming, et al. Research on weed detection algorithm in sugarcane field based on Faster R—CNN［J］. Journal of Chinese Agricultural Mechanization, 2024, 45（6）: 208-215.
［13］　彭明霞，夏俊芳，彭辉. 融合FPN的Faster R—CNN复杂背景下棉田杂草高效识别方法［J］. 农业工程学报，2019，35（20）: 202－209.
Peng Mingxia, Xia Junfang, Peng Hui. Efficient recognition of cotton and weed in field based on Faster R—CNN by integrating FPN ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35（20）: 202－209.
［14］　樊湘鹏，周建平，许燕，等. 基于优化Faster R—CNN的棉花苗期杂草识别与定位［J］. 农业机械学报, 2021, 52（5）: 26-34.
Fan Xiangpeng, Zhou Jianping, Xu Yan, et al. Identification and localization of weeds based on optimized Faster R—CNN in cotton seedling stage ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2020, 52（5）: 26-34.
［15］　李春明，逯杉婷，远松灵，等. 基于Faster R—CNN的除草机器人杂草识别算法［J］. 中国农机化学报，2019，40（12）:171-176.
Li Chunming, Lu Shanting, Yuan Songling, et al. Weed identification algorithm of weeding robot based on Faster R—CNN［J］. Journal of Chinese Agricultural Mechanization, 2019, 40（12）: 171-176.
［16］　薛金利，戴建国，赵庆展，等. 基于低空无人机影像和YOLOv3实现棉田杂草检测［J］. 石河子大学学报（自然科学版），2019, 37（1）: 21-27.
Xue Jinli, Dai Jianguo, Zhao Qingzhan, et al. Cotton field weed detection based on low-altitude drone image and YOLOv3 ［J］. Journal of Shihezi University (Natural Science)， 2019, 37（1）:21-27.
［17］　权龙哲，夏福霖，姜伟，等. 基于YOLOv4卷积神经网络的农田苗草识别研究［J］. 东北农业大学学报，2021，52（7）: 89-98.
Quan Longzhe, Xia Fulin, Jiang Wei, et al. Research on recognition of maize seedlings and weeds in maize mield based on YOLOv4 convolutional neural network ［J］. Journal of Northeast Agricultural University, 2021, 52（7）: 89-98.
［18］　尚钰莹，张倩如，宋怀波. 基于YOLOv5s的深度学习在自然场景苹果花朵检测中的应用［J］. 农业工程学报， 2022, 38（9）: 222-229.
Shang Yuying, Zhang Qianru, Song Huaibo. Application of deep learning using YOLOv5s to apple flower detection in natural scenes ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38（9）: 222-229.
［19］　孙丰刚，王云露，兰鹏，等. 基于改进YOLOv5s和迁移学习的苹果果实病害识别方法［J］. 农业工程学报. 2022，38（11）: 171-179.
Sun Fenggang, Wang Yunlu, Lan Peng, et al. Identification of apple fruit diseases using improved YOLOv5s and transfer learning ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38（11）: 171-179.
［20］　苏宇，肖志云，鲍鹏飞. 采用改进YOLOv5s 检测牧区牲畜［J］. 农业工程学报，2023，39（24）：165-176.
Su Yu, Xiao Zhiyun, Bao Pengfei. Livestock detection in pastoral areas using improved YOLOv5s ［J］. Transactions of the Chinese Society of Agricultural, 2023, 39（24）: 165-176.
［21］　朱圣, 邓继忠, 张亚莉, 等. 基于无人机低空遥感的水稻田间杂草分布图研究［J］. 华南农业大学学报, 2020, 41（6）: 67-74.
Zhu Sheng, Deng Jizhong, Zhang Yali, et al. Study on distribution map of weeds in rice field based on UAV remote sensing ［J］. Journal of South China Agricultural University, 2020, 41（6）: 67-74.
［22］　邓向武，梁松，许伊杰，等. 基于Subspace集成学习的稻田杂草识别研究［J］. 广东石油化工学院学报，2021，31（4）: 40-44.
Deng Xiangwu, Liang Song, Xu Yijie, et al. Weeds recognition in paddy fields based on subspace ensemble learning ［J］. Journal of Guangdong University of Petrochemical Technology, 2021, 31（4）: 40-44.
［23］　Ge Z, Liu S, Wang F, et al. YOLOX: Exceeding YOLO series in 2021 ［J］. arXiv preprint arXiv: 2107.08430, 2021.
［24］　陈禹，吴雪梅，张珍，等. 基于改进YOLOv5s的自然环境下茶叶病害识别方法［J］. 农业工程学报，2023, 39（24）: 185-194.
Chen Yu, Wu Xuemei, Zhang Zhen, et al. Method for identifying tea diseases in natural environment using improved YOLOv5s ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39（24）: 185-194.
［25］　章权兵，胡姗姗，舒文灿，等. 基于注意力机制金字塔网络的麦穗检测方法［J］. 农业机械学报, 2021, 52（11）: 253-262.
Zhang Quanbing, Hu Shanshan, Shu Wencan, et al. Wheat spikes detection method based on pyramidal network of attention mechanism ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52 （11）: 253-262.
［26］　杜甜甜，南新元，黄家興，等. 改进RegNet识别多种农作物病害受害程度［J］. 农业工程学报，2022，38（15）：150-158.
Du Tiantian, Nan Xinyuan, Huang Jiaxing, et al. Identifying the damage degree of various crop diseases using an improved RegNet ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38（15）: 150-158.
［27］　曾安，彭杰威，刘畅，等. 基于多尺度几何感知Transformer 的植物点云补全网络［J］. 农业工程学报, 2022, 38（4）：198-205.
Zeng An, Peng Jiewei, Liu Chang, et al. Plant point cloud completion network based on multi-scale geometry-aware point Transformer ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38（4）: 198-205.
［28］　Viola P A, Jones M J. Rapid object detection using a boosted cascade of simple features ［C］. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition, 2001: 511-518.
［29］　胡嘉沛，李震，黄河清，等. 采用改进YOLOv4—Tiny模型的柑橘木虱识别［J］. 农业工程学报, 2021, 37（17）：197-203.
Hu Jiapei, Li Zhen, Huang Heqing, et al. Citrus psyllid detection based on improved YOLOv4—Tiny model ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37（17）: 197-203.

[1]	段宗科, , 黄玉巧, , 李晓康, , 吕凤玉, , 古伟, , 丁立利, . 田间除草机器人机械结构设计与分析[J]. 中国农机化学报, 2025, 46(7): 21-27.
[2]	陈思, 李俊萩, 孔德肖, 张晴晖, 孙文朔. 基于YOLOv8n绿橙检测算法研究[J]. 中国农机化学报, 2025, 46(7): 145-152.
[3]	王真真, 彭强吉, 康建明, 王小瑜, 陈英凯. 果园机械除草发展分析及前景展望[J]. 中国农机化学报, 2025, 46(7): 190-197.
[4]	李思进, , 李金阳, , 张伟, . 基于改进YOLOv5s的大豆冠层叶片识别方法研究[J]. 中国农机化学报, 2025, 46(5): 100-105.
[5]	王靖, 姜文刚, 程耀, 钱伟. 基于Lab颜色空间改进U—Net的稻田杂草分割方法[J]. 中国农机化学报, 2025, 46(5): 148-154.
[6]	宋健, 郁志宏, 戚桂美, 解晶晶, 刘志兴, 王涛. 基于YOLOv8n+无人机图像的荒漠草原鼠洞检测[J]. 中国农机化学报, 2025, 46(5): 155-161.
[7]	陈佳豪, , 汪语哲, , 段晓东, , 梁凯华, . 基于改进YOLOv5s的青稞病虫害检测方法[J]. 中国农机化学报, 2025, 46(5): 162-168.
[8]	刘宇平, 刘程飞, 赵平伟. 基于DeepLabv3—Faster R—CNN的水稻叶片病害检测方法#br#[J]. 中国农机化学报, 2025, 46(4): 108-113.
[9]	丁展博, 徐健, 朱耀麟, 张勇进, 刘晨雨. 基于轻量化YOLO模型的花椒簇检测研究[J]. 中国农机化学报, 2025, 46(4): 120-125.
[10]	吕宗旺, 王甜甜, 孙福艳, 祝玉华, . 基于改进YOLOv8m的小麦仓储粮虫检测方法[J]. 中国农机化学报, 2025, 46(3): 108-114.
[11]	买买提·沙吾提, 阿尔庆·西力克. 基于改进YOLOv8的复杂果园环境下杏果的目标检测[J]. 中国农机化学报, 2025, 46(3): 246-252.
[12]	郭敬涛, 吕凤, 章慧婷, 杨彪, 刘大洋. 基于迁移学习与轻量化YOLOv5s的草莓目标检测方法[J]. 中国农机化学报, 2025, 46(3): 253-260.
[13]	许毓超, , 吴茜, , 张兵园, , , 周玲莉, , 任妮, , , 张美娜, , . 轻量级深度学习网络在农作物目标检测的应用进展[J]. 中国农机化学报, 2025, 46(3): 261-270.
[14]	罗柳茗, 李岩舟, 石美琦, 黄鑫, 陈汐 . 基于YOLOv8n的甘蔗杂草识别方法 [J]. 中国农机化学报, 2025, 46(2): 237-244.
[15]	夏子林, 张新洲, 王文波, 夏先飞, 陈兰, 顾寄南. 基于机器视觉的蚕豆荚高精度检测方法研究[J]. 中国农机化学报, 2025, 46(1): 157-163.