智能化农业装备学报（中英文）

融合MobileViT-CA机制与多任务学习的农业图像智能识别技术研究

预览PDF | 下载PDF

常清雅

智能化农业装备学报（中英文） | 2026,7(2): 153-160

收起

智能化农业装备学报（中英文） | 2026 , 7 (2) : 153 -160

融合MobileViT-CA机制与多任务学习的农业图像智能识别技术研究

全屏

[Author(id=1325194206574157982, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=qingya216280@126.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1325194206658044064, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, authorId=1325194206574157982, language=EN, stringName=Qingya Chang, firstName=Qingya, middleName=null, lastName=Chang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Automation Engineering, Xi'an Aeronautical Polytechnic Institute, Xi'an 710089, China, bio={"content":"

Chang Qingya, E-mail: qingya216280@126.com

"}, bioImg=null, bioContent=

Chang Qingya, E-mail: qingya216280@126.com

, aboutCorrespAuthor=null), CN=AuthorExt(id=1325194206788067489, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, authorId=1325194206574157982, language=CN, stringName=常清雅, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=西安航空职业技术学院自动化工程学院，陕西西安，710089, bio={"content":"

常清雅，女，1996年生，陕西渭南人，讲师，硕士；研究方向为电子技术、数据处理。E-mail：qingya216280@126.com

"}, bioImg=null, bioContent=

常清雅，女，1996年生，陕西渭南人，讲师，硕士；研究方向为电子技术、数据处理。E-mail：qingya216280@126.com

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1325194206490271898, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, xref=null, ext=[AuthorCompanyExt(id=1325194206498660507, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, companyId=1325194206490271898, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Automation Engineering, Xi'an Aeronautical Polytechnic Institute, Xi'an 710089, China), AuthorCompanyExt(id=1325194206511243420, tenantId=1146029695717560320, journalId=1319690209941618849, articleId=1325194205877903511, companyId=1325194206490271898, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=西安航空职业技术学院自动化工程学院，陕西西安，710089)])])]

常清雅

作者信息

西安航空职业技术学院自动化工程学院，陕西西安，710089

作者简介:

常清雅，女，1996年生，陕西渭南人，讲师，硕士；研究方向为电子技术、数据处理。E-mail：qingya216280@126.com

Research on agricultural image intelligent recognition technology integrating MobileViT-CA mechanism and multi-task learning

Qingya Chang

Affiliations

School of Automation Engineering, Xi'an Aeronautical Polytechnic Institute, Xi'an 710089, China

Chang Qingya, E-mail: qingya216280@126.com

出版时间: 2026-05-15 doi: 10.12398/j.issn.2096-7217.2026.02.014

文章导航

摘要

收起

针对当前农业图像识别任务中模型分散、多任务协同性弱、实时性与泛化能力不足的问题，研究提出一种基于深度学习与多任务融合的农业图像智能识别系统。该系统以“端—边—云”3级协同为核心架构，通过构建多任务联合学习框架，实现作物株高测量与病害识别的协同训练与推理。融合移动端卷积神经网络（mobile convolutional neural network，MobileNet）、视觉转换器（vision transformer，ViT）特征提取能力与坐标注意力（coordinate attention，CA）机制，在强化关键特征提取、抑制农田背景干扰的同时，大幅降低计算负担。试验结果表明，该系统在株高测量与病害识别两项核心任务中均取得显著成效。在株高测量方面，算法测量结果与人工实测值之间的决定系数R²最高达99.98%，在病害识别方面，改进后的轻量化模型平均检测精度达到98.76%，模型参数量仅为12.64 M，单张图像平均检测时间最低仅为9.1 ms，具备良好的嵌入式部署能力。此外，系统在复杂环境中仍保持稳定性能，株高测量偏差均在±2 cm范围内，可无缝嵌入无人机等低空监测装备，适配低空农业作业场景的实时数据处理需求。综上所述，研究系统兼具良好的可扩展性、实时性与鲁棒性，有效破解传统技术在多任务协同、轻量化部署与复杂环境适配中的瓶颈，为低空经济赋能下的智慧农业多维度图像识别提供一体化解决方案，助力精准农业规模化落地与农业生产提质增效。

关键词

农业图像识别 / 多任务学习 / 轻量化网络 / 注意力机制 / YOLO算法 / 低空农业监测

Abstract

收起

To address the limitations of current agricultural image recognition tasks, such as isolated models, weak multi-task synergy, insufficient real-time performance and generalization capability, this study develops an intelligent agricultural image recognition system based on deep learning and multi-task fusion. This system adopts a "device-edge-cloud" three-level collaborative architecture. By constructing a multi-task joint learning framework, the system enables concurrent training and collaborative inference for both crop height measurement and disease recognition. The system fuses the feature extraction capabilities of mobile convolutional neural networks (MobileNet) and visual transformers (ViT) with a coordinate attention (CA) mechanism; this design enhances key feature extraction, suppresses complex farmland background interference while significantly alleviating computational burden. Experimental results show that the system achieves remarkable performance across both core tasks. For crop height measurement, the determination coefficient R2 between the algorithm′s measurement results and the manual measured values reaches up to 99.98%. For disease recognition, the optimized lightweight model achieves an average detection accuracy of 98.76%, with only 12.64 M parameters. The average single-image inference latency is as low as 9.1 ms, demonstrating excellent potential for embedded deployment. Moreover, the system exhibits high stability in complex environments, maintaining measurement deviation of crop height within ±2 cm. This precision allows for seamless integration into low-altitude monitoring platforms, such as unmanned aerial vehicles (UAVs), to satisfy the real-time data processing requirements of low-altitude agricultural operations. In summary, the proposed system delivers excellent scalability, real-time efficiency, and robustness, effectively overcoming traditional bottlenecks in multi-task synergy, lightweight deployment, and adaptation to complex environments. It provides an integrated solution for multi-dimensional image recognition in smart agriculture empowered by the low-altitude economy, thereby facilitating the large-scale implementation of precision agriculture while improving agricultural production efficiency.

Key words

agricultural image recognition / multi-task learning / lightweight networks / attention mechanisms / YOLO algorithm / Agricultural monitoring at low altitudes

引用本文

常清雅. 融合MobileViT-CA机制与多任务学习的农业图像智能识别技术研究. 智能化农业装备学报（中英文）, 2026 , 7 (2) : 153 -160 . DOI: 10.12398/j.issn.2096-7217.2026.02.014

Qingya Chang. Research on agricultural image intelligent recognition technology integrating MobileViT-CA mechanism and multi-task learning[J]. Journal of Intelligent Agricultural Mechanization, 2026 , 7 (2) : 153 -160 . DOI: 10.12398/j.issn.2096-7217.2026.02.014

基金

收起

西安航空职业技术学院2025年度教育教学改革研究项目(25XHJG025)
陕西省教育科学“十四五”规划2025年度一般课题(SGH25Y4086)

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

[1]

何勇，王月影，何立文，等．低空经济政策和技术在农业农村的应用现状与前景[J]．农业工程学报，2025，41（8）：1-16.

He Yong, Wang Yueying, He Liwen, et al. Current status and prospects of low-altitude economy policies and technologies in agriculture and rural areas [J]. Transactions of the Chinese Society of Agricultural Engineering, 2025, 41(8): 1-16.

[2]

赵泽行，吴晓鹏，王怡馨，等．基于知识图谱的农作物病虫害问答系统研究[J]．智能化农业装备学报（中英文），2024，5（4）：39-50.

Zhao Zexing, Wu Xiaopeng, Wang Yixin, et al. Research on question answering system for crop diseases and pests based on knowledge graph [J]. Journal of Intelligent Agricultural Mechanization, 2024, 5(4): 39-50.

[3]

李家辉，李晗，牛永泽，等．精准农业技术对农户生产效率和家庭收入的协同影响及机制[J]．资源科学，2025，47（11）：2466-2479.

[4]

李康顺，杨振盛，江梓锋，等．基于改进YOLOX-Nano的农作物叶片病害检测与识别方法[J]．华南农业大学学报，2023，44（4）：593-603.

Li Kangshun, Yang Zhensheng, Jiang Zifeng, et al. Detection and identification of crop leaf diseases based on improved YOLOX-Nano [J]. Journal of South China Agricultural University, 2023, 44(4): 593-603.

[5]

陆娇娇，段峰峰．改进YOLOv5模型下农作物虫害图像识别算法仿真[J]．计算机仿真，2023，40（11）：156-160.

[6]

Fu X, Ma Q, Yang F, et al. Crop pest image recognition based on the improved ViT method [J]. Information Processing in Agriculture, 2024, 11(2): 249-259.

[7]

姜友谊，刘博伟，张成健，等．利用无人机多光谱影像的多品种玉米成熟度监测[J]．农业工程学报，2023，39（20）：84-91.

Jiang Youyi, Liu Bowei, Zhang Chengjian, et al. Monitoring the maturity of multiple varieties of corn using unmanned aerial vehicle multispectral images [J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(20): 84-91.

[8]

南德旺，李军营，梁虹，等．基于无人机图像和改进Yolov8的烟草植株株高检测[J]．计算机测量与控制，2025，33（7）：19-26.

Nan Dewang, Li Junying, Liang Hong, et al. Tobacco plant stem height detection based on UAV images and improved Yolov8 [J]. Computer Measurement and Control, 2025, 33(7): 19-26.

[9]

王娜，孟庆辉，魏宏飞．基于嵌入式S3C6410的采摘机器人自动导航控制系统[J]．农机化研究，2023，45（10）：182-186.

Wang Na, Meng Qinghui, Wei Hongfei. Design and experimental research of intelligent pipe vegetable planter[J]. Journal of Agricultural Mechanization Research, 2023, 45(10): 182-186.

[10]

卢中华，袁灿，马征，等．国内外中耕除草装备及其智能化技术现状及趋势[J]．智能化农业装备学报（中英文），2026，7（1）：129-135.

Lu Zhonghua, Yuan Can, Ma Zheng, et al. Current status and trends of intercultural weeding equipment and its intelligent technology at home and abroad [J]. Journal of Intelligent Agricultural Mechanization, 2026, 7 (1): 129-135.

[11]

王新蕾，王硕，翟嘉政，等．多任务联合学习下的复杂天气航拍图像目标检测算法[J]．计算机工程与应用，2025，61（2）：97-111.

Wang Xinlei, Wang Shuo, Zhai Jiazheng, et al. Object detection algorithm of aerial image in complex weather based on multi-task joint learning [J]. Computer Engineering and Applications, 2025, 61(2): 97-111.

[12]

张强，张灿智，曹恒，等．融合ViT与多尺度注意力的改进YOLOv8飞鸟识别算法[J]．科学技术与工程，2025，25（19）：8151-8157.

Zhang Qiang, Zhang Canzhi, Cao Heng, et al. Improved YOLOv8 bird recognition algorithm integrates ViT and multi-scale attention [J]. Science Technology and Engineering, 2025, 25(19): 8151-8157.

[13]

Zhang Q, Xu Y, Zhang J, et al. Vitaev2: Vision transformer advanced by exploring inductive bias for image recognition and beyond [J]. International Journal of Computer Vision, 2023, 131(5): 1141-1162..

[14]

Ammar L B, Gasmi K, Ltaifa I B. ViT-TB: ensemble learning based ViT model for tuberculosis recognition [J]. Cybernetics and Systems, 2024, 55(3): 634-653.

[15]

杨森森，张昊，兴陆，等．改进MobileViT网络识别轻量化田间杂草[J]．农业工程学报，2023，39（9）：152-160.

Yang Sensen, Zhang Hao, Xing Lu, et al. Light weight recognition of weeds in the field based on improved MobileViT network [J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39 (9): 152-160.

[16]

张欢，周毅，王克俭，等．基于MobileViT-PC-ASPP和迁移学习的果树害虫识别方法[J]．农业机械学报，2024，55（11）：57-67.

Zhang Huan, Zhou Yi, Wang Kejian, et al. Fruit tree pest identification method based on MobileViT-PC-ASPP and transfer learning [J]. Transactions of the Chinese Society for Agricultural Machinery, 2024, 55(11): 57-67.

2026年第7卷第2期

预览PDF | 下载PDF

引用本文

BibTeX

文章信息

doi: 10.12398/j.issn.2096-7217.2026.02.014

接收时间：2026-02-02
首发时间：2026-11-05
出版时间：2026-05-15

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2026-02-02
修回日期：2026-04-10

基金

2025 Education & Teaching Reform Research Project of Xi′an Aeronautical Polytechnic Institute(25XHJG025)

西安航空职业技术学院2025年度教育教学改革研究项目(25XHJG025)

2025 General Project of the 14th Five-Year Plan for Educational Science of Shaanxi Province(SGH25Y4086)

陕西省教育科学“十四五”规划2025年度一般课题(SGH25Y4086)

作者信息

西安航空职业技术学院自动化工程学院，陕西西安，710089

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/znhnyzbxb/CN/10.12398/j.issn.2096-7217.2026.02.014

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT