[1]王湲,姜新华,乌丹牧其尔,等.融合单通道注意力与蝴蝶优化算法的玉米叶片叶绿素和氮素含量预测[J].江苏农业学报,2026,42(06):1182-1190.[doi:doi:10.3969/j.issn.1000-4440.2026.06.010]
　WANG Yuan,JIANG Xinhua,Wudanmuqier,et al.Prediction of chlorophyll and nitrogen contents in maize leaves by fusing single-channel attention and butterfly optimization algorithm[J].,2026,42(06):1182-1190.[doi:doi:10.3969/j.issn.1000-4440.2026.06.010]
点击复制

融合单通道注意力与蝴蝶优化算法的玉米叶片叶绿素和氮素含量预测()

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
42
期数:
2026年06期
页码:
1182-1190
栏目:
农业信息工程
出版日期:
2026-06-30

文章信息/Info

Title:
Prediction of chlorophyll and nitrogen contents in maize leaves by fusing single-channel attention and butterfly optimization algorithm
作者:
王湲1; 2; 姜新华1; 2; 乌丹牧其尔1; 2; 张子汉1; 2; 付学良1; 2
(1.内蒙古农业大学计算机与信息工程学院,内蒙古呼和浩特010018;2.内蒙古自治区农牧业大数据研究与应用重点实验室,内蒙古呼和浩特010018)
Author(s):
WANG Yuan1; 2; JIANG Xinhua1; 2; Wudanmuqier1; 2; ZHANG Zihan1; 2; FU Xueliang1; 2
(1.College of Computer and Information Engineering, Inner Mongolia Agricultural University, Hohhot 010018, China;2.Inner Mongolia Autonomous Region Key Laboratory of Agricultural and Animal Husbandry Big Data Research and Application, Hohhot 010018, China)
关键词:
玉米叶片; 叶绿素含量; 氮素含量; RGB图像分析; 支持向量回归
Keywords:
maize leaves; chlorophyll content; nitrogen content; RGB image analysis; support vector regression
分类号:
S513文献标识码:A文章编号:1000-4440(2026)06-1182-09
DOI:
doi:10.3969/j.issn.1000-4440.2026.06.010
文献标志码:
A
摘要:
为满足精准农业对玉米叶片营养监测的快速无损检测需求,本研究提出一种融合图像分割、特征优化与参数寻优的机器学习预测方法。以智能手机采集的田间玉米叶片RGB图像为数据源,采用分割任意模型（SAM）技术实现叶片区域高精度分割,并基于R(红色)、G(绿色)、B(蓝色)三通道构建42种颜色特征。在建模阶段,将单通道注意力机制(AM)与蝴蝶优化算法(BOA)嵌入支持向量回归(SVR)框架,构建BOA-AM-SVR协同预测模型。其中,AM通过梯度下降迭代更新特征权重,实现对高价值颜色特征的自适应强化;BOA模拟蝴蝶种群觅食行为,以验证集决定系数(R2)作为适应度函数,优化SVR的核参数与惩罚系数。在叶绿素含量预测方面,BOA-AM-SVR模型的R2为0.860 7,平均绝对误差(MAE)为2.317 2;在氮素含量预测方面,其R2为0.854 0,MAE为0.755 9。本研究结果为田间作物营养实时诊断与精准施肥提供了技术支持。
Abstract:
To meet the demand for rapid and non-destructive detection in precision agriculture for nutrient monitoring of maize leaves, this study proposed a machine learning prediction method fusing image segmentation, feature optimization, and parameter optimization. Using RGB images of field maize leaves collected by smartphones as the data source, the Segment Anything Model (SAM) technique was adopted to achieve high-precision segmentation of leaf regions. Based on the R (red), G (green), and B (blue) channels, 42 color features were constructed. In the modeling stage, a single-channel attention mechanism (AM) and the butterfly optimization algorithm (BOA) were embedded into the support vector regression (SVR) framework to construct a BOA-AM-SVR collaborative prediction model. Among them, AM adaptively enhanced high-value color features by iteratively updating feature weights through gradient descent. BOA simulated the foraging behavior of butterfly populations and took the coefficient of determination (R2) of the validation set as the fitness function to optimize the kernel parameters and penalty coefficients of SVR. For chlorophyll content prediction, the R2 of the BOA-AM-SVR model was 0.860 7 and the mean absolute error (MAE) was 2.317 2. For nitrogen content prediction, the R2 was 0.854 0 and the MAE was 0.755 9. This study provides technical support for real-time diagnosis of field crop nutrition and precise fertilization.

参考文献/References:

[1]ARNON D I. Copper enzymes in isolated chloroplasts. Polyphenoloxidase in Beta vulgaris[J]. Plant physiology,1949,24(1):1.
[2]范志影,刘庆生,张萍. 用凯氏法和杜马斯法测定植物样品中的全氮[J]. 现代科学仪器,2007(1):46-47.
[3]DUTTA GUPTA S, IBARAKI Y, PATTANAYAK A K. Development of a digital image analysis method for real-time estimation of chlorophyll content in micropropagated potato plants[J]. Plant Biotechnology Reports,2013,7(1):91-97.
[4]薛宪法. 基于图像和光谱技术的玉米叶片叶绿素与氮素含量检测[D]. 杨凌:西北农林科技大学,2018.
[5]李丽,程灵. 基于RGB模型建立快速检测番茄叶片叶绿素含量的方法[J]. 分子植物育种,2021,19(20):6906-6909.
[6]张立周,王殿武,张玉铭,等. 数字图像技术在夏玉米氮素营养诊断中的应用[J]. 中国生态农业学报,2010,18(6):1340-1344.
[7]白金顺,曹卫东,熊静,等. 应用数码相机进行绿肥翻压后春玉米氮素营养诊断和产量预测[J]. 光谱学与光谱分析,2013,33(12):3334-3338.
[8]苏月,温福东,高鑫彤,等. 基于支持向量回归模型预测中国包虫病发病例数[J]. 哈尔滨医科大学学报,2025,59(2):211-215.
[9]杨镇远,喻红芹,杨飞,等. 机器学习在成纱质量预测模型中的研究进展[J]. 上海纺织科技,2025,53(8):14-19.
[10]YE L, YOU D H, YIN X G, et al. An improved fault-location method for distribution system using wavelets and support vector regression[J]. International Journal of Electrical Power & Energy Systems, 2014,55:467-472.
[11]黄亚坤. 基于多源数据的工作面顶板涌水量预测研究[D]. 廊坊:华北科技学院,2025.
[12]ARORA S, SINGH S. Butterfly optimization algorithm:a novel approach for global optimization[J]. Soft Computing,2019,23(3):715-734.
[13]KARABOGA D, BASTURK B. A powerful and efficient algorithm for numerical function optimization:artificial bee colony (ABC) algorithm[J]. Journal of Global Optimization,2007,39(3):459-471.
[14]张铃,张钹. 遗传算法机理的研究[J]. 软件学报,2000(7):945-952.
[15]VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[EB/OL]. （2017-06-12）
[2024-10-08]. https://doi.org/10.48550/arXiv.1706.03762.
[16]WANG X L, GIRSHICK R, GUPTA A, et al. Non-local neural networks[EB/OL]. (2018-04-13)
[2025-09-10]. https://arxiv.org/abs/1711.07971.
[17]DOS SANTOS COELHO L, MARIANI V C. Improved differential evolution algorithms for handling economic dispatch optimization with generator constraints[J]. Energy Conversion and Management,2007,48(5):1631-1639.
[18]宋词. 计算机视觉中注意力机制的策略选择与优化研究[D]. 哈尔滨:黑龙江大学,2023.
[19]李学艺,王相凯,黄金凤,等. 基于双通道信号驱动的CA-Transformer轴承故障诊断方法[J]. 组合机床与自动化加工技术,2025(7):121-127.
[20]凌琪涵,孔发明,宁强,等. 基于无人机多光谱影像的水稻氮营养监测[J]. 农业工程学报,2023,39(13):160-170.
[21]夏晨真,姜艳艳,张星宇,等. 基于无人机高光谱影像的黑土区玉米农田土壤有机质估算[J]. 光谱学与光谱分析,2023,43(8):2617-2626.
[22]张新生,李忆楠,陈章政,等. 基于GBDT-XGBoost的西北地区制造业碳排放预测研究[J]. 环境科学与技术,2024,47(12):131-143.
[23]韩松岳,王明宇,端军红,等. 基于发散学习麻雀算法的大模型部署架构及优化方法[J]. 军事运筹与评估,2024,39(6):61-68.
[24]刁兴春,刘艺,曹建军,等. 多目标蚁群优化研究综述[J]. 计算机科学,2017,44(10):7-13,25.
[25]张琪,顾腾达,任宇辰,等. 多策略改进的蝴蝶优化算法[J]. 计算机工程与科学,2025,47(7):1312-1320.
[26]HOU Y M, JIA S Y, LUN X M, et al. GCNs-net:a graph convolutional neural network approach for decoding time-resolved eeg motor imagery signals[J]. IEEE Transactions on Neural Networks and Learning Systems,2022,35(6):7312-7323.

相似文献/References:

[1]安飞飞,简纯平,杨龙,等.木薯幼苗叶绿素含量及光合特性对盐胁迫的响应[J].江苏农业学报,2015,(03):500.[doi:10.3969/j.issn.1000-4440.2015.03.006]
　AN Fei-fei,JIAN Chun-ping,YANG Long,et al.Chlorophyll contents and photosynthetic characteristics of cassava seedlings in response to NaCl stress[J].,2015,(06):500.[doi:10.3969/j.issn.1000-4440.2015.03.006]
[2]孙彦坤,陈睿,李静,等.不同降雨年型下反枝苋和大豆光合特征的比较[J].江苏农业学报,2019,(03):554.[doi:doi:10.3969/j.issn.1000-4440.2019.03.008]
　SUN Yan-kun,CHEN Rui,LI Jing,et al.Comparison of photosynthetic characteristics between Amaranthus retroexus and Glycine max under different annual rainfall pattern[J].,2019,(06):554.[doi:doi:10.3969/j.issn.1000-4440.2019.03.008]
[3]郭松,常庆瑞,郑智康,等.基于无人机高光谱影像的玉米叶绿素含量估测[J].江苏农业学报,2022,38(04):976.[doi:doi:10.3969/j.issn.1000-4440.2022.04.014]
　GUO Song,CHANG Qing-rui,ZHENG Zhi-kang,et al.Estimation of maize chlorophyll content based on unmanned aerial vehicle (UAV) hyperspectral images[J].,2022,38(06):976.[doi:doi:10.3969/j.issn.1000-4440.2022.04.014]
[4]王宇,汪泓,肖玖军,等.基于MCC-GAPLS-PLSR的辣椒叶绿素含量高光谱定量反演[J].江苏农业学报,2024,(05):865.[doi:doi:10.3969/j.issn.1000-4440.2024.05.011]
　WANG Yu,WANG Hong,XIAO Jiujun,et al.Hyperspectral quantitative inversion of chlorophyll content in pepper based on MCC-GAPLS-PLSR[J].,2024,(06):865.[doi:doi:10.3969/j.issn.1000-4440.2024.05.011]
[5]唐国强,刘梦云,蒋丹垚,等.基于分数阶微分的猕猴桃叶片叶绿素含量估算[J].江苏农业学报,2025,(02):335.[doi:doi:10.3969/j.issn.1000-4440.2025.02.014]
　TANG Guoqiang,LIU Mengyun,JIANG Danyao,et al.Estimation of kiwifruit leaf chlorophyll content based on fractional-order differential processing[J].,2025,(06):335.[doi:doi:10.3969/j.issn.1000-4440.2025.02.014]
[6]郭思思,任春梅,李硕,等.二氢卟吩铁调控玉米幼苗生长及耐盐性[J].江苏农业学报,2025,(03):485.[doi:doi:10.3969/j.issn.1000-4440.2025.03.007]
　GUO Sisi,REN Chunmei,LI Shuo,et al.Effect of iron dihydroporphine on the growth and salt resistance of maize seedlings[J].,2025,(06):485.[doi:doi:10.3969/j.issn.1000-4440.2025.03.007]

备注/Memo

备注/Memo:
收稿日期:2025-08-18基金项目:内蒙古自治区科技重大专项(2021ZD0003);内蒙古自然科学基金项目(2023LHMS06017、2024LHMS06021);内蒙古自然科学基金重点项目(2025ZD012);内蒙古自治区揭榜挂帅项目(2025KJTW0026);内蒙古自治区重点项目(2025KYPT0076)；国家自然科学基金项目(62061037、31960494);内蒙古自治区科技攻关计划项目(2020GG0169)作者简介:王湲(2000-),女,内蒙古呼和浩特人,硕士研究生,研究方向为模式识别与智能信息处理。(E-mail)wy112828@163.com通讯作者:姜新华,(E-mail)jiangxh@imau.edu.cn
更新日期/Last Update: 2026-07-15