Pub Date : 2026-03-15Epub Date: 2026-01-27DOI: 10.1016/j.compag.2026.111456
Qixin Kang , Guozhong Zhang , Chao Ji , Zhuangzhuang Zhao , Liming Chen , Lingxiao Lan
<div><div><em>Allium chinense</em> is a specialty crop valued for its distinctive flavor and medicinal properties, with substantial exports to Japan, South Korea, and international markets. Despite its economic importance, cultivation remains labor-intensive due to strict agronomic demands for precise seed orientation. A key obstacle to mechanization is collision-induced bouncing during sowing, which compromises placement quality. This study investigates the effects of seed-discharge parameters and seedbed geometry on directional sowing by analyzing seed–soil collision mechanisms. DEM (Discrete Element Method) models based on HM-JKR (Hertz-Mindlin with JKR) model were constructed, with parameters calibrated via P-B (Plackett-Burman) screening and B-B (Box-Behnken) optimization. Targeted at an AOR (angle of repose) of 41.67°, the optimal parameters were determined as restitution coefficient (<em>R<sub>e</sub></em>) = 0.051, rolling friction coefficient (<em>C<sub>r</sub></em>) = 0.027, and surface energy (<em>K<sub>i</sub></em>) = 0.180 J/m<sup>2</sup> between soil and seeds, yielding a relative error of 1.8 %. Single-factor simulation experiments were conducted to analyze seed-discharge height (<em>h<sub>d</sub></em>), velocity (<em>v<sub>d</sub></em>) and posture (<em>p<sub>d</sub></em>) on flat and V-shaped seedbeds from an impact-energy perspective. Uniform-design experiments were conducted to develop multivariate regression prediction models, which were then validated by simulations and bench tests. Results indicate that on flat seedbeds, <em>h<sub>d</sub></em>, <em>v<sub>d</sub></em>, and <em>p<sub>d</sub></em> significantly influence maximum bounce height (<em>h<sub>b</sub></em>) and rebound velocity (<em>v<sub>b</sub></em>), whereas planar displacement (<em>e<sub>p</sub></em>) is primarily governed by <em>h<sub>d</sub></em> and <em>v<sub>d</sub></em>. On V-shaped seedbeds, <em>v<sub>d</sub></em> exhibits a highly significant effect on <em>v<sub>b</sub></em> but no significant effect on <em>h<sub>b</sub></em> or <em>e<sub>p</sub></em>. As <em>v<sub>d</sub></em> increases, <em>h<sub>b</sub></em> and <em>v<sub>b</sub></em> follow a “first increase, then decrease” trend. In contrast, seed bounce generally intensifies on flat seedbeds. Validation tests showed no significant differences between model predictions, simulations, and bench tests (paired t-tests, p > 0.05), supporting model reliability. Under sandy loam conditions, using lower seed-discharge kinetic energy combined with matched furrow geometry and seed-discharge posture effectively mitigates bounce and drift of <em>Allium chinense</em> seeds. The optimized parameters are <em>h<sub>d</sub></em> = 180 mm and <em>v<sub>d</sub></em> = 0.36 m/s, with seeds oriented facing the furrow and aligned either parallel or perpendicular to the furrow axis. These findings provide actionable guidelines for mechanized oriented sowing systems targeting <em>Allium chinense</em> and analogous bulbous
中国Allium chinense是一种特色作物，因其独特的风味和药用价值而受到重视，大量出口到日本，韩国和国际市场。尽管具有重要的经济意义，但由于严格的农艺要求精确的种子定位，种植仍然是劳动密集型的。机械化的一个关键障碍是播种过程中碰撞引起的弹跳，这会影响播种质量。通过对种子-土壤碰撞机理的分析，探讨了种子流量参数和苗床几何形状对定向播种的影响。基于HM-JKR （Hertz-Mindlin with JKR）模型构建DEM （Discrete Element Method）模型，通过P-B （Plackett-Burman）筛选和B-B （Box-Behnken）优化对参数进行标定。以休止角（AOR）为41.67°为目标，确定了土壤与种子间的最优参数：恢复系数(Re) = 0.051，滚动摩擦系数(Cr) = 0.027，表面能(Ki) = 0.180 J/m2，相对误差为1.8%。通过单因素模拟实验，从冲击能的角度分析了平床和v型床的排种高度（hd）、排种速度（vd）和排种姿态（pd）。采用均匀设计实验建立多元回归预测模型，并通过模拟和台架试验对模型进行验证。结果表明，在平坦的苗床上，hd、vd和pd显著影响最大反弹高度（hb）和反弹速度（vb），而平面位移（ep）主要受hd和vd的影响。在v型苗床上，vd对vb的影响极显著，对hb和ep的影响不显著。随着vd的增加，hb和vb呈现先增加后减少的趋势。相反，在平坦的苗床上，种子反弹通常会加剧。验证检验显示，模型预测、模拟和台架检验之间无显著差异（配对t检验，p > 0.05），支持模型可靠性。在砂壤土条件下，采用较低的排种动能，配合匹配的畦型和排种姿态，可有效缓解葱种子的弹跳和漂移。优化后的参数为hd = 180 mm, vd = 0.36 m/s，种子朝向犁沟，平行或垂直于犁沟轴线。这些发现为中国葱及类似球茎作物的机械化定向播种系统提供了可操作的指导。
{"title":"Coupled effects of seed-discharge parameters and seedbed geometry on Allium chinense seeds–soil bounce","authors":"Qixin Kang , Guozhong Zhang , Chao Ji , Zhuangzhuang Zhao , Liming Chen , Lingxiao Lan","doi":"10.1016/j.compag.2026.111456","DOIUrl":"10.1016/j.compag.2026.111456","url":null,"abstract":"<div><div><em>Allium chinense</em> is a specialty crop valued for its distinctive flavor and medicinal properties, with substantial exports to Japan, South Korea, and international markets. Despite its economic importance, cultivation remains labor-intensive due to strict agronomic demands for precise seed orientation. A key obstacle to mechanization is collision-induced bouncing during sowing, which compromises placement quality. This study investigates the effects of seed-discharge parameters and seedbed geometry on directional sowing by analyzing seed–soil collision mechanisms. DEM (Discrete Element Method) models based on HM-JKR (Hertz-Mindlin with JKR) model were constructed, with parameters calibrated via P-B (Plackett-Burman) screening and B-B (Box-Behnken) optimization. Targeted at an AOR (angle of repose) of 41.67°, the optimal parameters were determined as restitution coefficient (<em>R<sub>e</sub></em>) = 0.051, rolling friction coefficient (<em>C<sub>r</sub></em>) = 0.027, and surface energy (<em>K<sub>i</sub></em>) = 0.180 J/m<sup>2</sup> between soil and seeds, yielding a relative error of 1.8 %. Single-factor simulation experiments were conducted to analyze seed-discharge height (<em>h<sub>d</sub></em>), velocity (<em>v<sub>d</sub></em>) and posture (<em>p<sub>d</sub></em>) on flat and V-shaped seedbeds from an impact-energy perspective. Uniform-design experiments were conducted to develop multivariate regression prediction models, which were then validated by simulations and bench tests. Results indicate that on flat seedbeds, <em>h<sub>d</sub></em>, <em>v<sub>d</sub></em>, and <em>p<sub>d</sub></em> significantly influence maximum bounce height (<em>h<sub>b</sub></em>) and rebound velocity (<em>v<sub>b</sub></em>), whereas planar displacement (<em>e<sub>p</sub></em>) is primarily governed by <em>h<sub>d</sub></em> and <em>v<sub>d</sub></em>. On V-shaped seedbeds, <em>v<sub>d</sub></em> exhibits a highly significant effect on <em>v<sub>b</sub></em> but no significant effect on <em>h<sub>b</sub></em> or <em>e<sub>p</sub></em>. As <em>v<sub>d</sub></em> increases, <em>h<sub>b</sub></em> and <em>v<sub>b</sub></em> follow a “first increase, then decrease” trend. In contrast, seed bounce generally intensifies on flat seedbeds. Validation tests showed no significant differences between model predictions, simulations, and bench tests (paired t-tests, p > 0.05), supporting model reliability. Under sandy loam conditions, using lower seed-discharge kinetic energy combined with matched furrow geometry and seed-discharge posture effectively mitigates bounce and drift of <em>Allium chinense</em> seeds. The optimized parameters are <em>h<sub>d</sub></em> = 180 mm and <em>v<sub>d</sub></em> = 0.36 m/s, with seeds oriented facing the furrow and aligned either parallel or perpendicular to the furrow axis. These findings provide actionable guidelines for mechanized oriented sowing systems targeting <em>Allium chinense</em> and analogous bulbous ","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111456"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146079878","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-01-23DOI: 10.1016/j.compag.2026.111465
Hao Wang , Yixue Liu , Bin Sun , Juncheng Ma , Chao Liang , Xiao Yang , Renli Qi , Chaoyuan Wang
Floor fouling monitoring in pig facilities is essential for early disease detection and environmental hygiene management, as diarrheal feces indicates digestive disorders while manure accumulation directly impacts animal health and welfare. Current manual inspection methods are labor-intensive and subjective, while existing computer vision approaches suffer from unstable color features under varying lighting conditions and misclassification of background textures as fouling patterns. To address these challenges, we propose FreCANet, a frequency-aware deep learning framework that achieves multi-level fouling classification through hierarchical visual interference suppression. The method integrates three key innovations: Mask R-CNN preprocessing that eliminates pig body occlusion (improving detection recall by up to 17.21%), Frequency Dynamic Convolution that separates manure contamination features from environmental noise across different frequency bands, and Efficient Channel Attention embedded within residual connections for selective feature enhancement. Using a comprehensive dataset of 25,228 images covering seven fouling categories across the complete growth cycle, FreCANet achieved 88.31% accuracy and 0.8679 F1-Score, outperforming ResNet-152 by 2.44% and 2.93% respectively. Diarrheal feces detection reached 95.9% precision on slatted floors and 89.3% recall on solid floors, enabling reliable early warning for digestive health issues. The four-level manure contamination classification achieved 77.2–87.4% precision across fouling gradients from clean to severely soiled conditions. These results demonstrate FreCANet’s effectiveness in transforming subjective manual inspection into quantitative pen hygiene assessment for precision livestock farming applications.
{"title":"Frequency-aware deep learning for diarrheal feces and floor fouling monitoring in pig pens","authors":"Hao Wang , Yixue Liu , Bin Sun , Juncheng Ma , Chao Liang , Xiao Yang , Renli Qi , Chaoyuan Wang","doi":"10.1016/j.compag.2026.111465","DOIUrl":"10.1016/j.compag.2026.111465","url":null,"abstract":"<div><div>Floor fouling monitoring in pig facilities is essential for early disease detection and environmental hygiene management, as diarrheal feces indicates digestive disorders while manure accumulation directly impacts animal health and welfare. Current manual inspection methods are labor-intensive and subjective, while existing computer vision approaches suffer from unstable color features under varying lighting conditions and misclassification of background textures as fouling patterns. To address these challenges, we propose FreCANet, a frequency-aware deep learning framework that achieves multi-level fouling classification through hierarchical visual interference suppression. The method integrates three key innovations: Mask R-CNN preprocessing that eliminates pig body occlusion (improving detection recall by up to 17.21%), Frequency Dynamic Convolution that separates manure contamination features from environmental noise across different frequency bands, and Efficient Channel Attention embedded within residual connections for selective feature enhancement. Using a comprehensive dataset of 25,228 images covering seven fouling categories across the complete growth cycle, FreCANet achieved 88.31% accuracy and 0.8679 F1-Score, outperforming ResNet-152 by 2.44% and 2.93% respectively. Diarrheal feces detection reached 95.9% precision on slatted floors and 89.3% recall on solid floors, enabling reliable early warning for digestive health issues. The four-level manure contamination classification achieved 77.2–87.4% precision across fouling gradients from clean to severely soiled conditions. These results demonstrate FreCANet’s effectiveness in transforming subjective manual inspection into quantitative pen hygiene assessment for precision livestock farming applications.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111465"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To reduce the draught force and energy consumption of the mouldboard plough, a biomimetic plough body surface (BPBS) was designed with the structures of pangolin scales and shark placoid scales as biomimetic coupling elements. A soil-mouldboard plough interaction model was constructed by EDEM and verified for accuracy through laboratory soil bin experiments under sandy loam conditions. The verified model was employed for optimising the biomimetic coupling plough body surface. Single-factor, steepest ascent and Box-Behnken experiments were designed to investigate the transverse spacing, longitudinal spacing, installation angle, scale length, and biomimetic structure height of the BPBS with the draught force and the soil disturbance area used as indicators. Results illustrated both errors of simulated and measured draught forces and soil disturbance areas were less than 5.8%. The transverse spacing, longitudinal spacing, installation angle, and biomimetic structure height of BPBS influenced the two indexes significantly. The optimal parameters from Box-Behnken response surface experiments were transverse spacing of 23 mm, longitudinal spacing of 15 mm, installation angle of 30°, and biomimetic structure height of 14 mm. BPBS with the optimal parameters reduced the draught force by 13.06% and increased the soil disturbance area by 5.29% compared with the conventional mouldboard plough. The errors of predicted values were small compared with the simulated. The design of the BPBS achieved the research objective of reducing the draught force without affecting the ploughing quality.
{"title":"Biomimetic design and research of plough body surface for reducing draught force of a single large-width mouldboard plough","authors":"Hao Zhou, Zhiyu Qin, Xuezhen Wang, Shengsheng Wang, Kangtai Li, Guoliang Deng, Haibo Yan","doi":"10.1016/j.compag.2026.111461","DOIUrl":"10.1016/j.compag.2026.111461","url":null,"abstract":"<div><div>To reduce the draught force and energy consumption of the mouldboard plough, a biomimetic plough body surface (BPBS) was designed with the structures of pangolin scales and shark placoid scales as biomimetic coupling elements.<!--> <!-->A soil-mouldboard plough interaction model was constructed by EDEM and verified for accuracy through laboratory soil bin experiments under sandy loam conditions. The verified model was employed for optimising the biomimetic coupling plough body surface.<!--> <!-->Single-factor, steepest ascent and Box-Behnken experiments were designed to investigate the transverse spacing, longitudinal spacing, installation angle, scale length, and biomimetic structure height of the BPBS with the draught force and the soil disturbance area used as indicators. Results illustrated both errors of simulated and measured draught forces and soil disturbance areas were less than 5.8%. The transverse spacing, longitudinal spacing, installation angle, and biomimetic structure height of BPBS influenced the two indexes significantly. The optimal parameters from Box-Behnken response surface experiments were transverse spacing of 23 mm, longitudinal spacing of 15 mm, installation angle of 30°, and biomimetic structure height of 14 mm. BPBS with the optimal parameters reduced the draught force by 13.06% and increased the soil disturbance area by 5.29% compared with the conventional mouldboard plough. The errors of predicted values were small compared with the simulated. The design of the BPBS achieved the research objective of reducing the draught force without affecting the ploughing quality.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111461"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-01-21DOI: 10.1016/j.compag.2026.111442
Whai-En Chen, Shih-Che Lin, Hsin-Hung Cho
Piglet mortality is a significant issue on pig farms. Often, piglets do not receive proper care after birth or are crushed by the sow, leading to the death of 2 to 4 piglets during each farrowing process. This represents a substantial loss for pig farmers. To address these issues, this paper presents the development of the PigTalk2.0 system, which is based on the Artificial Internet of Things (AIoT) and image recognition technology, aimed at preventing piglet deaths caused by crushing by the sow. Since collecting large-scale agricultural datasets is costly and labor-intensive, Since collecting large-scale agricultural datasets is costly, this study proposes a hybrid augmentation strategy. We combine conventional techniques (e.g., translation, flipping) with Generative AI-based data augmentation to effectively expand limited datasets and enhance model robustness, providing synthetic diversity that strengthens model training. PigTalk2.0 uses image recognition to monitor the posture of the sow. To prevent the sow from accidentally crushing the piglets when she suddenly lies down, the system activates an air blower to gently push the piglets away when the sow stands up. Additionally, PigTalk2.0 can detect when a sow is about to farrow for the first time, allowing the system to notify the farm administrator to check the piglets’ health and provide necessary care immediately after birth. The results demonstrate that introducing consumer IoT technology into family-run agriculture or medium- to large-scale enterprise-run pig farms can significantly improve productivity.
{"title":"Generative data augmentation for AIoT-based smart pig farming: The PigTalk 2.0 system","authors":"Whai-En Chen, Shih-Che Lin, Hsin-Hung Cho","doi":"10.1016/j.compag.2026.111442","DOIUrl":"10.1016/j.compag.2026.111442","url":null,"abstract":"<div><div>Piglet mortality is a significant issue on pig farms. Often, piglets do not receive proper care after birth or are crushed by the sow, leading to the death of 2 to 4 piglets during each farrowing process. This represents a substantial loss for pig farmers. To address these issues, this paper presents the development of the PigTalk2.0 system, which is based on the Artificial Internet of Things (AIoT) and image recognition technology, aimed at preventing piglet deaths caused by crushing by the sow. Since collecting large-scale agricultural datasets is costly and labor-intensive, Since collecting large-scale agricultural datasets is costly, this study proposes a hybrid augmentation strategy. We combine conventional techniques (e.g., translation, flipping) with Generative AI-based data augmentation to effectively expand limited datasets and enhance model robustness, providing synthetic diversity that strengthens model training. PigTalk2.0 uses image recognition to monitor the posture of the sow. To prevent the sow from accidentally crushing the piglets when she suddenly lies down, the system activates an air blower to gently push the piglets away when the sow stands up. Additionally, PigTalk2.0 can detect when a sow is about to farrow for the first time, allowing the system to notify the farm administrator to check the piglets’ health and provide necessary care immediately after birth. The results demonstrate that introducing consumer IoT technology into family-run agriculture or medium- to large-scale enterprise-run pig farms can significantly improve productivity.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111442"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-02-06DOI: 10.1016/j.compag.2026.111527
Jie Jiang , Zhaopeng Fu , Jiayi Zhang , Jinpeng Yang , Wanping Fang , Qiang Cao , Yongchao Tian , Yan Zhu , Weixing Cao , Xiaojun Liu
Management zone (MZ) mapping is the first step in precision nitrogen management (PNM) to improve crop management efficiency and optimize resource input. Satellite remote sensing is a suitable tool for collecting crop growth information to guide decision-making over large areas. This study aimed to develop a PNM strategy utilizing satellite image-based MZ mapping to improve crop nitrogen use efficiency and reduce energy costs. Four wheat field experiments were conducted from 2020 to 2022 in Xinghua County, China. Sentinel-2 satellite imagery was employed to delineate wheat growth MZs using the Fuzzy K-means clustering method prior to the nitrogen (N) topdressing period. Subsequently, a management zone-accumulated N deficit-based N topdressing algorithm (MZ-ANDA) was developed to calculate N recommendation rates and evaluate the effects of variable N application in comparison to individual ANDA and conventional farmer fixed N application practices. The results demonstrated that wheat growth MZs could be effectively delineated using multi-temporal Sentinel-2 satellite spectral images before the N topdressing period. The MZ-ANDA strategy showed comparable results to ANDA in single wheat fields. However, when considering the comprehensive benefits of variable N topdressing at large area farms, the MZ-ANDA strategy reduced energy input by 2.22–26.54 GJ, economic input by $511.91–2639.08, and CO2 emissions by 0.31–2.05 tons across three farms ranging from 57.21 to 283.78 ha. Compared to conventional farmer fixed N practices, the MZ-ANDA reduced the N application rates with 12.67%–37.00% and increased the partial factor productivity with 9.47%–19.81% while maintained a high grain yield, meanwhile, the MZ-ANDA improved the net profit (9.23–139.48 $ ha−1), energy use efficiency (4.10%–10.17%) and energy productivity (5.06%–11.66%) while reduced the CO2 emission (2.95%–10.00%). In summary, the developed N management strategy using satellite images-based MZ map can significantly increase management efficiency and optimize resource input for in-season wheat management at the county scale. Future research could focus on refining the MZ-ANDA algorithm for different crops and regions to further enhance N management efficiency.
{"title":"Developing a nitrogen management strategy for winter wheat to enhance economic profit and energy savings using satellite-based management zone mapping at the county scale","authors":"Jie Jiang , Zhaopeng Fu , Jiayi Zhang , Jinpeng Yang , Wanping Fang , Qiang Cao , Yongchao Tian , Yan Zhu , Weixing Cao , Xiaojun Liu","doi":"10.1016/j.compag.2026.111527","DOIUrl":"10.1016/j.compag.2026.111527","url":null,"abstract":"<div><div>Management zone (MZ) mapping is the first step in precision nitrogen management (PNM) to improve crop management efficiency and optimize resource input. Satellite remote sensing is a suitable tool for collecting crop growth information to guide decision-making over large areas. This study aimed to develop a PNM strategy utilizing satellite image-based MZ mapping to improve crop nitrogen use efficiency and reduce energy costs. Four wheat field experiments were conducted from 2020 to 2022 in Xinghua County, China. Sentinel-2 satellite imagery was employed to delineate wheat growth MZs using the Fuzzy K-means clustering method prior to the nitrogen (N) topdressing period. Subsequently, a management zone-accumulated N deficit-based N topdressing algorithm (MZ-ANDA) was developed to calculate N recommendation rates and evaluate the effects of variable N application in comparison to individual ANDA and conventional farmer fixed N application practices. The results demonstrated that wheat growth MZs could be effectively delineated using multi-temporal Sentinel-2 satellite spectral images before the N topdressing period. The MZ-ANDA strategy showed comparable results to ANDA in single wheat fields. However, when considering the comprehensive benefits of variable N topdressing at large area farms, the MZ-ANDA strategy reduced energy input by 2.22–26.54 GJ, economic input by $511.91–2639.08, and CO<sub>2</sub> emissions by 0.31–2.05 tons across three farms ranging from 57.21 to 283.78 ha. Compared to conventional farmer fixed N practices, the MZ-ANDA reduced the N application rates with 12.67%–37.00% and increased the partial factor productivity with 9.47%–19.81% while maintained a high grain yield, meanwhile, the MZ-ANDA improved the net profit (9.23–139.48 $ ha<sup>−1</sup>), energy use efficiency (4.10%–10.17%) and energy productivity (5.06%–11.66%) while reduced the CO<sub>2</sub> emission (2.95%–10.00%). In summary, the developed N management strategy using satellite images-based MZ map can significantly increase management efficiency and optimize resource input for in-season wheat management at the county scale. Future research could focus on refining the MZ-ANDA algorithm for different crops and regions to further enhance N management efficiency.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111527"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Identifying the spatial distribution and degree of soil salinity accurately and promptly is crucial for reclaiming salt-induced wasteland. It provides decision support for rapidly delineating the center of salt patches, zoning salinized areas, and rationally planning reclamation regions. This study investigated the potential of deriving Salinization-Feature Enhanced (SFE) spectral indices from multispectral (MS) band combinations for estimating field soil salinity content (SSC) in the Luohui Canal Irrigation District, Shaanxi China. The sensitivity of the SFE indices and conventional spectral indices (SIs) to SSC was analyzed using the Pearson correlation coefficient and the Variable Iterative Space Shrinkage Approach (VISSA). We assessed the accuracy of estimating surface SSC during the seedling stage under complex soil salinization environments, analyzing the effects of varying spatial resolutions and sampling densities. Extreme Learning Machine (ELM) and Random Forest Classifier (RFC) were employed to invert surface SSC and generate a digital soil salinity map. The results indicated that the SFE (|r| = 0.58) indices showed a stronger correlation with SSC compared with the SIs (|r| = 0.26), indicating superior performance in surface SSC inversion. The incorporation of environmental covariates enhanced the model’s inversion accuracy and stability. The sample dataset was divided into 70% training and 30% test subsets using stratified sampling. Surface SSC prediction models were developed using SFE indices and SIs with both the ELM and RFC algorithms. The ELM_SFE (accuracy = 0.68) and RFC_SIs (accuracy = 0.69) models achieved the best predictive accuracy. The ELM_SFE and RFC_SFE models demonstrated the best inversion performance in severely and moderately salt-induced wasteland, respectively. This study provides valuable insights for precision agricultural management in salt-induced wasteland reclamation.
{"title":"Precision mapping of soil salinity in reclaiming salt-induced wasteland with UAV multispectral images and machine learning","authors":"Haisong Yao , Tibin Zhang , Songling Chen , Yuxin Kuang , Yu Cheng , Qing Liang , Hao Feng , Gaoxiang Zhou , Kadambot H.M. Siddique","doi":"10.1016/j.compag.2026.111532","DOIUrl":"10.1016/j.compag.2026.111532","url":null,"abstract":"<div><div>Identifying the spatial distribution and degree of soil salinity accurately and promptly is crucial for reclaiming salt-induced wasteland. It provides decision support for rapidly delineating the center of salt patches, zoning salinized areas, and rationally planning reclamation regions. This study investigated the potential of deriving Salinization-Feature Enhanced (SFE) spectral indices from multispectral (MS) band combinations for estimating field soil salinity content (SSC) in the Luohui Canal Irrigation District, Shaanxi China. The sensitivity of the SFE indices and conventional spectral indices (SIs) to SSC was analyzed using the Pearson correlation coefficient and the Variable Iterative Space Shrinkage Approach (VISSA). We assessed the accuracy of estimating surface SSC during the seedling stage under complex soil salinization environments, analyzing the effects of varying spatial resolutions and sampling densities. Extreme Learning Machine (ELM) and Random Forest Classifier (RFC) were employed to invert surface SSC and generate a digital soil salinity map. The results indicated that the SFE (|r| = 0.58) indices showed a stronger correlation with SSC compared with the SIs (|r| = 0.26), indicating superior performance in surface SSC inversion. The incorporation of environmental covariates enhanced the model’s inversion accuracy and stability. The sample dataset was divided into 70% training and 30% test subsets using stratified sampling. Surface SSC prediction models were developed using SFE indices and SIs with both the ELM and RFC algorithms. The ELM_SFE (accuracy = 0.68) and RFC_SIs (accuracy = 0.69) models achieved the best predictive accuracy. The ELM_SFE and RFC_SFE models demonstrated the best inversion performance in severely and moderately salt-induced wasteland, respectively. This study provides valuable insights for precision agricultural management in salt-induced wasteland reclamation.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111532"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173948","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate crop mapping is conducive to optimizing agricultural production, improving resource utilization efficiency, as well as supporting precision agriculture management and environmental monitoring. However, crop mapping often relies on a large amount of reference data as labels. Moreover, the generalization and scalability of approaches are challenging issues due to the influence of geographic locations, climatic conditions, and crop characteristics. Thereby, this study proposed the Domain Adaptation with Multiple Classifiers (MCDA) model that enables crop mapping in target domains without reference data, which consists of a feature extractor and three classifiers. Initially, the data from the source domain was trained using the feature extractor and individual classifier. Next, the feature extractor was fixed and two classifiers were trained. The decision boundaries of classifiers were maximized by calculating discrepancies between different classification results in target domain. Finally, the two classifiers were fixed and the feature extractor was trained to map the data from the source domain and target domain into a better feature space. In this study, three sets of experiments were designed to validate the stability and generalization of MCDA model by selecting different study areas in four countries to classify typical crops such as maize and winter wheat, rice, and soybeans using time series Vegetation Indices (VIs). Comparing the 1DCNN and LSTM models, as well as the four UDA models, the experimental results indicate that the MCDA model outperforms the other models in most transfer cases. With the appropriate selection of crop features, the method proposed in this study can achieve more accurate crop mapping effectively identifying crop types and field boundaries. The implementation code of our method has been publicly available at https://github.com/abbyxuchanghong-cmd/MCDA.
准确的作物测图有利于优化农业生产，提高资源利用效率，支持精准农业管理和环境监测。然而，作物映射通常依赖于大量的参考数据作为标签。此外，由于地理位置、气候条件和作物特性的影响，方法的泛化和可扩展性是具有挑战性的问题。为此，本研究提出了一种无需参考数据即可实现目标域作物映射的多分类器域适应模型（Domain Adaptation with Multiple Classifiers， MCDA），该模型由一个特征提取器和三个分类器组成。首先，使用特征提取器和个体分类器对源域的数据进行训练。然后，固定特征提取器，训练两个分类器。通过计算目标域内不同分类结果之间的差异，最大化分类器的决策边界。最后，固定两个分类器，训练特征提取器将源域和目标域的数据映射到更好的特征空间。为了验证MCDA模型的稳定性和泛化性，本研究设计了3组实验，选取4个国家的不同研究区域，利用时间序列植被指数（VIs）对玉米、冬小麦、水稻和大豆等典型作物进行分类。对比1DCNN和LSTM模型以及4种UDA模型，实验结果表明MCDA模型在大多数迁移情况下都优于其他模型。通过对作物特征的适当选择，本研究方法可以实现更精确的作物作图，有效识别作物类型和田间边界。我们的方法的实现代码可以在https://github.com/abbyxuchanghong-cmd/MCDA上公开获得。
{"title":"MCDA: a novel domain adaptation with multiple classifiers for crop mapping","authors":"Changhong Xu , Maofang Gao , Yuanwei Chen , Jingwen Yan","doi":"10.1016/j.compag.2026.111450","DOIUrl":"10.1016/j.compag.2026.111450","url":null,"abstract":"<div><div>Accurate crop mapping is conducive to optimizing agricultural production, improving resource utilization efficiency, as well as supporting precision agriculture management and environmental monitoring. However, crop mapping often relies on a large amount of reference data as labels. Moreover, the generalization and scalability of approaches are challenging issues due to the influence of geographic locations, climatic conditions, and crop characteristics. Thereby, this study proposed the Domain Adaptation with Multiple Classifiers (MCDA) model that enables crop mapping in target domains without reference data, which consists of a feature extractor and three classifiers. Initially, the data from the source domain was trained using the feature extractor and individual classifier. Next, the feature extractor was fixed and two classifiers were trained. The decision boundaries of classifiers were maximized by calculating discrepancies between different classification results in target domain. Finally, the two classifiers were fixed and the feature extractor was trained to map the data from the source domain and target domain into a better feature space. In this study, three sets of experiments were designed to validate the stability and generalization of MCDA model by selecting different study areas in four countries to classify typical crops such as maize and winter wheat, rice, and soybeans using time series Vegetation Indices (VIs). Comparing the 1DCNN and LSTM models, as well as the four UDA models, the experimental results indicate that the MCDA model outperforms the other models in most transfer cases. With the appropriate selection of crop features, the method proposed in this study can achieve more accurate crop mapping effectively identifying crop types and field boundaries. The implementation code of our method has been publicly available at <span><span>https://github.com/abbyxuchanghong-cmd/MCDA</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111450"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146174052","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-02-05DOI: 10.1016/j.compag.2026.111528
Biao Zhang, Qiancheng Liu, Jishan Wen, Guangyi Liu
For continuous row sowing, automatic cane-seed metering devices face the challenge of efficiently sorting stalk-seeds from a chaotically stacked state into an orderly discrete arrangement, but the geometric variability of such irregular stalks and their multiple interactions with flow channels complicate the controllability of seed posture and flow flux. To address this, a novel sorting approach based on metered mechanical energy charge-release was proposed to adaptively regulate the orientation and flux of high-mass-energy rod-particle flow. The diversion chains of ordering and discretization were thus sequentially integrated. Combining interactive DEM-MBD simulations with bench experimental behavior, the dynamic structure–activity relationship of energetic drive-constraint sequence was investigated as particle arrangement progressively transformed into ordered supply and quantitative discharge. This revealed the normalized contact mechanics mechanism for directional alignment and individual disengagement. The synergistic utilization of forked rolling, unbalanced collision and linear acceleration was demonstrated to produce a straightening flow effect, and the redundancy elimination effect achievable by the V-shaped turning lifting flow was elucidated. Logistics response models were developed to reflect the impact of various diversion factors on sequential sorting capacities. With the optimal regulation path coordinated for energy-flux, the results showed that arrangement orderliness and discrete uniformity were improved by 50.3% and 86.3%, respectively, implying more stable control over transfer posture consistency and discharge accuracy. Such improvements in sorting quality while maintaining efficiency provided direct evidence for the controllable transformation of irregular stalk-seed arrangement states.
{"title":"A sorting approach capable of ordering and discretizing stacked stalk-seeds by mechanical energy metering","authors":"Biao Zhang, Qiancheng Liu, Jishan Wen, Guangyi Liu","doi":"10.1016/j.compag.2026.111528","DOIUrl":"10.1016/j.compag.2026.111528","url":null,"abstract":"<div><div>For continuous row sowing, automatic cane-seed metering devices face the challenge of efficiently sorting stalk-seeds from a chaotically stacked state into an orderly discrete arrangement, but the geometric variability of such irregular stalks and their multiple interactions with flow channels complicate the controllability of seed posture and flow flux. To address this, a novel sorting approach based on metered mechanical energy charge-release was proposed to adaptively regulate the orientation and flux of high-mass-energy rod-particle flow. The diversion chains of ordering and discretization were thus sequentially integrated. Combining interactive DEM-MBD simulations with bench experimental behavior, the dynamic structure–activity relationship of energetic drive-constraint sequence was investigated as particle arrangement progressively transformed into ordered supply and quantitative discharge. This revealed the normalized contact mechanics mechanism for directional alignment and individual disengagement. The synergistic utilization of forked rolling, unbalanced collision and linear acceleration was demonstrated to produce a straightening flow effect, and the redundancy elimination effect achievable by the V-shaped turning lifting flow was elucidated. Logistics response models were developed to reflect the impact of various diversion factors on sequential sorting capacities. With the optimal regulation path coordinated for energy-flux, the results showed that arrangement orderliness and discrete uniformity were improved by 50.3% and 86.3%, respectively, implying more stable control over transfer posture consistency and discharge accuracy. Such improvements in sorting quality while maintaining efficiency provided direct evidence for the controllable transformation of irregular stalk-seed arrangement states.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111528"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173988","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-02-04DOI: 10.1016/j.compag.2026.111522
Sally Deborah Pereira da Silva , Vinicius Richter , Norton Borges Junior , Regiane Aparecida Ferreira , Gustavo Vedooto Ferreira , Telmo Jorge Carneiro Amado , Luan Pierre Pott , Lucio de Paula Amaral
Root malformation disorder (RMD) is an abiotic stress that compromises the early development of Eucalyptus saligna plantations in Brazil, reducing growth, nutrient uptake, and canopy vigor. Despite its operational relevance, scalable tools for objective field detection remain limited. Remote sensing using unmanned aerial vehicles (UAVs) combined with deep learning offers a promising alternative for identifying early physiological stress. The objectives of this work were: (i) to characterize the biophysical attributes of plants affected by RMD; and (ii) to evaluate the feasibility of a deep learning–based approach to map different plant health conditions of E. saligna at the stand scale. Multispectral data from a RedEdge-MX sensor (Blue, Green, Red, Red-edge, NIR) were collected over an 11 ha, six-month-old E. saligna stand in Southern Brazil. Field measurements included plant height, diameter at breast height (DBH), chlorophyll content, and leaf nutrient concentrations. Ten vegetation indices (VIs) were computed, and Random Forest (Gini importance) identified two key predictors: the Canopy Chlorophyll Content Index (CCCI) and the Plant Senescing Reflectance Index (PSRI). U-Net++ was trained to classify four classes: healthy, unhealthy, dead plants, and soil/residues. RMD-affected trees showed significant reductions in height, DBH, chlorophyll content, and nutrient concentrations. The combination CCCI + PSRI yielded the best discrimination of unhealthy plants (precision = 98.75%, recall = 92.94%, F1-score = 95.76%), with an overall accuracy of 98.77%. Applied to the full stand, 93.41% of trees were classified as healthy, 3.70% as unhealthy, and 2.90% as dead. These findings demonstrate that UAV multispectral imagery integrated with U-Net++ enables accurate, low-cost detection of RMD-related stress, supporting early silvicultural decision-making and routine plantation monitoring.
{"title":"Remote detection of root malformation disorder in Eucalyptus saligna using UAV multispectral imagery and U-Net++","authors":"Sally Deborah Pereira da Silva , Vinicius Richter , Norton Borges Junior , Regiane Aparecida Ferreira , Gustavo Vedooto Ferreira , Telmo Jorge Carneiro Amado , Luan Pierre Pott , Lucio de Paula Amaral","doi":"10.1016/j.compag.2026.111522","DOIUrl":"10.1016/j.compag.2026.111522","url":null,"abstract":"<div><div>Root malformation disorder (RMD) is an abiotic stress that compromises the early development of <em>Eucalyptus saligna</em> plantations in Brazil, reducing growth, nutrient uptake, and canopy vigor. Despite its operational relevance, scalable tools for objective field detection remain limited. Remote sensing using unmanned aerial vehicles (UAVs) combined with deep learning offers a promising alternative for identifying early physiological stress. The objectives of this work were: (i) to characterize the biophysical attributes of plants affected by RMD; and (ii) to evaluate the feasibility of a deep learning–based approach to map different plant health conditions of <em>E. saligna</em> at the stand scale. Multispectral data from a RedEdge-MX sensor (Blue, Green, Red, Red-edge, NIR) were collected over an 11 ha, six-month-old <em>E. saligna</em> stand in Southern Brazil. Field measurements included plant height, diameter at breast height (DBH), chlorophyll content, and leaf nutrient concentrations. Ten vegetation indices (VIs) were computed, and Random Forest (Gini importance) identified two key predictors: the Canopy Chlorophyll Content Index (CCCI) and the Plant Senescing Reflectance Index (PSRI). U-Net++ was trained to classify four classes: healthy, unhealthy, dead plants, and soil/residues. RMD-affected trees showed significant reductions in height, DBH, chlorophyll content, and nutrient concentrations. The combination CCCI + PSRI yielded the best discrimination of unhealthy plants (precision = 98.75%, recall = 92.94%, F1-score = 95.76%), with an overall accuracy of 98.77%. Applied to the full stand, 93.41% of trees were classified as healthy, 3.70% as unhealthy, and 2.90% as dead. These findings demonstrate that UAV multispectral imagery integrated with U-Net++ enables accurate, low-cost detection of RMD-related stress, supporting early silvicultural decision-making and routine plantation monitoring.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111522"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-15Epub Date: 2026-01-28DOI: 10.1016/j.compag.2026.111443
Chang Xu , Lei Zhao , Haojie Wen , Yiding Zhang , Lipo Wang , Lingxian Zhang
Efficient management and precise monitoring are essential for the sustainable control of crop diseases and pests. Traditional unimodal methods exhibit reduced reliability due to data gaps and environmental fluctuations. Multimodal artificial intelligence (AI) offers a promising alternative by integrating complementary data sources and enhancing robustness and adaptability. However, a comprehensive synthesis connecting multimodal AI with multi-scale disease and pest management is still lacking. Based on 950 publications from the past decade reflecting a 31.7% annual growth rate over the past five years, this review examines the evolution of AI-driven research and compares unimodal and multimodal approaches by summarizing major data modalities, fusion strategies, and modeling techniques. Deep learning emerges as the most widely used class of AI methods, and quantitative evidence indicates that multimodal systems achieve approximately 3–48.9% higher diagnostic accuracy than unimodal models. Evidence from 27 studies demonstrates the effectiveness of multimodal fusion across imaging, spectral, environmental, and sensor-based datasets. Building upon these findings, we propose a novel three-level management framework comprising point-level diagnosis, area-scale monitoring, and spatiotemporal forecasting, clarifying how multimodal AI strengthens each task. We further highlight the role of Plant Electronic Medical Records (PEMRs) and outline a conceptual virtual plant clinic to support continuous, data-driven crop health services. Finally, this review identifies key directions including advanced fusion strategies, lightweight and interpretable models, digital twin integration, and scalable decision-support systems, which are essential for intelligent and sustainable crop disease and pest management.
{"title":"Intelligent management of crop diseases and pests in multiscale and multimodal complex scenarios: Technologies, applications, and prospects","authors":"Chang Xu , Lei Zhao , Haojie Wen , Yiding Zhang , Lipo Wang , Lingxian Zhang","doi":"10.1016/j.compag.2026.111443","DOIUrl":"10.1016/j.compag.2026.111443","url":null,"abstract":"<div><div>Efficient management and precise monitoring are essential for the sustainable control of crop diseases and pests. Traditional unimodal methods exhibit reduced reliability due to data gaps and environmental fluctuations. Multimodal artificial intelligence (AI) offers a promising alternative by integrating complementary data sources and enhancing robustness and adaptability. However, a comprehensive synthesis connecting multimodal AI with multi-scale disease and pest management is still lacking. Based on 950 publications from the past decade reflecting a 31.7% annual growth rate over the past five years, this review examines the evolution of AI-driven research and compares unimodal and multimodal approaches by summarizing major data modalities, fusion strategies, and modeling techniques. Deep learning emerges as the most widely used class of AI methods, and quantitative evidence indicates that multimodal systems achieve approximately 3–48.9% higher diagnostic accuracy than unimodal models. Evidence from 27 studies demonstrates the effectiveness of multimodal fusion across imaging, spectral, environmental, and sensor-based datasets. Building upon these findings, we propose a novel three-level management framework comprising point-level diagnosis, area-scale monitoring, and spatiotemporal forecasting, clarifying how multimodal AI strengthens each task. We further highlight the role of Plant Electronic Medical Records (PEMRs) and outline a conceptual virtual plant clinic to support continuous, data-driven crop health services. Finally, this review identifies key directions including advanced fusion strategies, lightweight and interpretable models, digital twin integration, and scalable decision-support systems, which are essential for intelligent and sustainable crop disease and pest management.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"244 ","pages":"Article 111443"},"PeriodicalIF":8.9,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：