Pub Date : 2025-01-06DOI: 10.1016/j.aiia.2025.01.001
Xue Xia , Ning Zhang , Zhibin Guan , Xin Chai , Shixin Ma , Xiujuan Chai , Tan Sun
Aggressive behavior among piglets is considered a harmful social contact. Monitoring weaned piglets with intense aggressive behaviors is paramount for pig breeding management. This study introduced a novel hybrid model, PAB-Mamba-YOLO, integrating the principles of Mamba and YOLO for efficient visual detection of weaned piglets' aggressive behaviors, including climbing body, nose hitting, biting tail and biting ear. Within the proposed model, a novel CSPVSS module, which integrated the Cross Stage Partial (CSP) structure with the Visual State Space Model (VSSM), has been developed. This module was adeptly integrated into the Neck part of the network, where it harnessed convolutional capabilities for local feature extraction and leveraged the visual state space to reveal long-distance dependencies. The model exhibited sound performance in detecting aggressive behaviors, with an average precision (AP) of 0.976 for climbing body, 0.994 for nose hitting, 0.977 for biting tail and 0.994 for biting ear. The mean average precision (mAP) of 0.985 reflected the model's overall effectiveness in detecting all classes of aggressive behaviors. The model achieved a detection speed FPS of 69 f/s, with model complexity measured by 7.2 G floating-point operations (GFLOPs) and parameters (Params) of 2.63 million. Comparative experiments with existing prevailing models confirmed the superiority of the proposed model. This work is expected to contribute a glimmer of fresh ideas and inspiration to the research field of precision breeding and behavioral analysis of animals.
{"title":"PAB-Mamba-YOLO: VSSM assists in YOLO for aggressive behavior detection among weaned piglets","authors":"Xue Xia , Ning Zhang , Zhibin Guan , Xin Chai , Shixin Ma , Xiujuan Chai , Tan Sun","doi":"10.1016/j.aiia.2025.01.001","DOIUrl":"10.1016/j.aiia.2025.01.001","url":null,"abstract":"<div><div>Aggressive behavior among piglets is considered a harmful social contact. Monitoring weaned piglets with intense aggressive behaviors is paramount for pig breeding management. This study introduced a novel hybrid model, PAB-Mamba-YOLO, integrating the principles of Mamba and YOLO for efficient visual detection of weaned piglets' aggressive behaviors, including climbing body, nose hitting, biting tail and biting ear. Within the proposed model, a novel CSPVSS module, which integrated the Cross Stage Partial (CSP) structure with the Visual State Space Model (VSSM), has been developed. This module was adeptly integrated into the Neck part of the network, where it harnessed convolutional capabilities for local feature extraction and leveraged the visual state space to reveal long-distance dependencies. The model exhibited sound performance in detecting aggressive behaviors, with an average precision (AP) of 0.976 for climbing body, 0.994 for nose hitting, 0.977 for biting tail and 0.994 for biting ear. The mean average precision (mAP) of 0.985 reflected the model's overall effectiveness in detecting all classes of aggressive behaviors. The model achieved a detection speed FPS of 69 f/s, with model complexity measured by 7.2 G floating-point operations (GFLOPs) and parameters (Params) of 2.63 million. Comparative experiments with existing prevailing models confirmed the superiority of the proposed model. This work is expected to contribute a glimmer of fresh ideas and inspiration to the research field of precision breeding and behavioral analysis of animals.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 1","pages":"Pages 52-66"},"PeriodicalIF":8.2,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-03DOI: 10.1016/j.aiia.2024.12.004
Xionghai Chen , Fei Yuan , Syed Tahir Ata-Ul-Karim , Xiaojun Liu , Yongchao Tian , Yan Zhu , Weixing Cao , Qiang Cao
Soil organic matter (SOM) is a key metric for assessing soil quality and crop yield potential. It plays a vital role in maintaining the ecological balance environment and promoting sustainable farming practices. This review examines the evolving trends in remote sensing (RS)-based SOM monitoring by analyzing 739 scholarly publications from the Web of Science database from 2003 to 2023 using a bibliometric approach. The study reveals that research on RS-based SOM monitoring has entered a rapid growth phase since 2018, with China and the United States as the main contributors and an extensive international cooperation network. In model construction, high frequency covariates such as soil pH, precipitation, temperature, and topography significantly improved the prediction accuracy. Data preprocessing methods such as Standard Normal Variables (SNV), Principal Component Analysis (PCA), and Multiple Scattering Correction (MSC) enhanced data consistency. Traditional statistical models are gradually being replaced by nonlinear machine learning and deep learning methods (CNN, XGBoost, andStacking), which are particularly good at handling complex high-dimensional data. Regional spectral libraries (OzSoil and AfSIS) excel in local accuracy, while global spectral libraries (ISRIC and LUCAS) are more suitable for cross-region modeling, and the migration learning technique effectively improves the model generalization ability in low data regions. Integrated models (CNN-LSTM and GAN) have significant advantages in capturing the spatial and temporal dynamics of SOMs, and uncertainty quantification methods (Bayesian inference, Monte Carlo simulation) enhance the reliability of the models in multi-source data and data-scarce scenarios. Future research should focus on further optimization of multi-source data fusion and uncertainty quantification to promote the development and applicability of RS-based SOM monitoring techniques for precision soil management and sustainable agriculture.
土壤有机质（SOM）是评价土壤质量和作物产量潜力的重要指标。它在维持生态平衡环境和促进可持续耕作方式方面发挥着至关重要的作用。本文采用文献计量学方法，分析了2003 - 2023年Web of Science数据库中739篇学术论文，探讨了基于遥感（RS）的SOM监测的发展趋势。研究表明，自2018年以来，基于rs的SOM监测研究进入快速增长阶段，以中美两国为主要贡献者，形成了广泛的国际合作网络。在模型构建中，土壤pH、降水、温度、地形等高频协变量显著提高了预测精度。采用标准正态变量（Standard Normal Variables， SNV）、主成分分析（Principal Component Analysis， PCA）和多重散射校正（Multiple Scattering Correction， MSC）等数据预处理方法增强了数据的一致性。传统的统计模型正逐渐被非线性机器学习和深度学习方法（CNN、XGBoost和stacking）所取代，这些方法特别擅长处理复杂的高维数据。区域谱库（OzSoil和AfSIS）具有较好的局部精度，而全局谱库（ISRIC和LUCAS）更适合跨区域建模，迁移学习技术有效提高了低数据区的模型泛化能力。集成模型（CNN-LSTM和GAN）在捕获SOMs时空动态方面具有显著优势，不确定性量化方法（贝叶斯推理、蒙特卡罗模拟）增强了模型在多源数据和数据稀缺场景下的可靠性。未来的研究应进一步优化多源数据融合和不确定性量化，以促进基于rs的SOM监测技术在土壤精准管理和可持续农业中的发展和应用。
{"title":"A bibliometric analysis of research on remote sensing-based monitoring of soil organic matter conducted between 2003 and 2023","authors":"Xionghai Chen , Fei Yuan , Syed Tahir Ata-Ul-Karim , Xiaojun Liu , Yongchao Tian , Yan Zhu , Weixing Cao , Qiang Cao","doi":"10.1016/j.aiia.2024.12.004","DOIUrl":"10.1016/j.aiia.2024.12.004","url":null,"abstract":"<div><div>Soil organic matter (SOM) is a key metric for assessing soil quality and crop yield potential. It plays a vital role in maintaining the ecological balance environment and promoting sustainable farming practices. This review examines the evolving trends in remote sensing (<em>RS</em>)-based SOM monitoring by analyzing 739 scholarly publications from the Web of Science database from 2003 to 2023 using a bibliometric approach. The study reveals that research on RS-based SOM monitoring has entered a rapid growth phase since 2018, with China and the United States as the main contributors and an extensive international cooperation network. In model construction, high frequency covariates such as soil pH, precipitation, temperature, and topography significantly improved the prediction accuracy. Data preprocessing methods such as Standard Normal Variables (SNV), Principal Component Analysis (PCA), and Multiple Scattering Correction (MSC) enhanced data consistency. Traditional statistical models are gradually being replaced by nonlinear machine learning and deep learning methods (CNN, XGBoost, andStacking), which are particularly good at handling complex high-dimensional data. Regional spectral libraries (OzSoil and AfSIS) excel in local accuracy, while global spectral libraries (ISRIC and LUCAS) are more suitable for cross-region modeling, and the migration learning technique effectively improves the model generalization ability in low data regions. Integrated models (CNN-LSTM and GAN) have significant advantages in capturing the spatial and temporal dynamics of SOMs, and uncertainty quantification methods (Bayesian inference, Monte Carlo simulation) enhance the reliability of the models in multi-source data and data-scarce scenarios. Future research should focus on further optimization of multi-source data fusion and uncertainty quantification to promote the development and applicability of RS-based SOM monitoring techniques for precision soil management and sustainable agriculture.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 1","pages":"Pages 26-38"},"PeriodicalIF":8.2,"publicationDate":"2025-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149330","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-02DOI: 10.1016/j.aiia.2024.12.005
John Olamofe , Ram Ray , Xishuang Dong , Lijun Qian
In this study, we examined plant health prediction through the Normalized Difference Vegetation Index (NDVI) calculated from satellite image derived reflectance values in the near-infrared and red spectra. The problem is formulated as a temporal data prediction problem. Using MODIS/Terra Vegetation Indices 16-Day L3 Global 250 m SIN Grid V061 dataset, we designed and implemented Reservoir Computing (RC) models and transformer-based models including pretrained language model, and compared the prediction performance of these models to traditional machine learning and deep learning methods such as Nonlinear Regression, Decision Tree, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) network, and DLinear. It is observed that the DLinear/LSTM model showed exceptional predictive accuracy, while the pretrained RC model significantly enhanced traditional RC model forecasts. Additionally, Frozen Pretrained Transformer (FPT), a pretrained language model, showed superior performance in predicting specific NDVI values (most often peak or lowest NDVI), suggesting its effectiveness in precise temporal predictions. Furthermore, transformer-based models, specifically PatchTST and FPT, demonstrated substantial mean squared error reductions, particularly in limited data scenarios (1 %, 5 %, 15 % and 50 % sample sizes), indicating their robustness in precise NDVI temporal predictions when data is limited. The findings in this study demonstrated the effectiveness of emerging machine learning techniques such as reservoir computing and pretrained language model for remote sensing and their contributions in precision agriculture.
在这项研究中，我们通过卫星图像近红外和红色光谱的反射率值计算的归一化植被指数（NDVI）来检验植物健康预测。该问题被表述为一个时间数据预测问题。利用MODIS/Terra植被指数16天L3全球250 m SIN网格V061数据集，设计并实现了水库计算（RC）模型和基于变压器的预训练语言模型，并将这些模型的预测性能与传统的机器学习和深度学习方法（如非线性回归、决策树、卷积神经网络（CNN）、长短期记忆（LSTM）网络和DLinear）进行了比较。结果表明，DLinear/LSTM模型具有较好的预测精度，而预训练后的RC模型显著提高了传统RC模型的预测精度。此外，Frozen Pretrained Transformer （FPT）是一种预训练语言模型，在预测特定的NDVI值（通常是峰值或最低NDVI）方面表现优异，表明其在精确时间预测方面的有效性。此外，基于变压器的模型，特别是PatchTST和FPT，显示出显著的均方误差降低，特别是在有限的数据场景下（1%、5%、15%和50%的样本量），表明它们在数据有限时精确的NDVI时间预测中的鲁棒性。本研究的发现证明了水库计算和遥感预训练语言模型等新兴机器学习技术的有效性及其在精准农业中的贡献。
{"title":"Normalized difference vegetation index prediction using reservoir computing and pretrained language models","authors":"John Olamofe , Ram Ray , Xishuang Dong , Lijun Qian","doi":"10.1016/j.aiia.2024.12.005","DOIUrl":"10.1016/j.aiia.2024.12.005","url":null,"abstract":"<div><div>In this study, we examined plant health prediction through the Normalized Difference Vegetation Index (NDVI) calculated from satellite image derived reflectance values in the near-infrared and red spectra. The problem is formulated as a temporal data prediction problem. Using MODIS/Terra Vegetation Indices 16-Day L3 Global 250 m SIN Grid V061 dataset, we designed and implemented Reservoir Computing (RC) models and transformer-based models including pretrained language model, and compared the prediction performance of these models to traditional machine learning and deep learning methods such as Nonlinear Regression, Decision Tree, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) network, and DLinear. It is observed that the DLinear/LSTM model showed exceptional predictive accuracy, while the pretrained RC model significantly enhanced traditional RC model forecasts. Additionally, Frozen Pretrained Transformer (FPT), a pretrained language model, showed superior performance in predicting specific NDVI values (most often peak or lowest NDVI), suggesting its effectiveness in precise temporal predictions. Furthermore, transformer-based models, specifically PatchTST and FPT, demonstrated substantial mean squared error reductions, particularly in limited data scenarios (1 %, 5 %, 15 % and 50 % sample sizes), indicating their robustness in precise NDVI temporal predictions when data is limited. The findings in this study demonstrated the effectiveness of emerging machine learning techniques such as reservoir computing and pretrained language model for remote sensing and their contributions in precision agriculture.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 1","pages":"Pages 116-129"},"PeriodicalIF":8.2,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-24DOI: 10.1016/j.aiia.2024.12.001
Weiwei Wang , Wenbing Shi , Ce Liu , Yuwei Wang , Lu Liu , Liqing Chen
With advancements in agricultural technology, the full mechanization of rice straw wheat planting has been achieved. However, issues such as missed seeding, uneven row spacing, and poor uniformity of row replenishment often arise due to wheel slippage in wheeled wheat seeders. These problems manual replanting after emergence, reducing efficiency and increasing labor costs. To address these challenges, a speed-adaptive wheat seeding control system based on speed radar was developed. This system comprises a pneumatic wheat seeding device, an automatic speed-following control system, a human-machine interface, and a stepper motor. Leveraging an embedded controller, the system dynamically adjusts motor speed based on real-time forward speed to ensure precise seeding. Using fuzzy PID control, the system dynamically adjusts motor speed, achieving row spacing consistency below 3.9 % and seeding stability within 1.3 %, even at varying speeds. This system addresses critical challenges in precision agriculture, enhancing planting efficiency and reducing labor costs. This innovation enhances planting efficiency, reduces labor costs, and ensures adaptability to varying tractor speeds, meeting the precision requirements of wheat planting.
{"title":"Development of automatic wheat seeding quantity control system based on Doppler radar speed measurement","authors":"Weiwei Wang , Wenbing Shi , Ce Liu , Yuwei Wang , Lu Liu , Liqing Chen","doi":"10.1016/j.aiia.2024.12.001","DOIUrl":"10.1016/j.aiia.2024.12.001","url":null,"abstract":"<div><div>With advancements in agricultural technology, the full mechanization of rice straw wheat planting has been achieved. However, issues such as missed seeding, uneven row spacing, and poor uniformity of row replenishment often arise due to wheel slippage in wheeled wheat seeders. These problems manual replanting after emergence, reducing efficiency and increasing labor costs. To address these challenges, a speed-adaptive wheat seeding control system based on speed radar was developed. This system comprises a pneumatic wheat seeding device, an automatic speed-following control system, a human-machine interface, and a stepper motor. Leveraging an embedded controller, the system dynamically adjusts motor speed based on real-time forward speed to ensure precise seeding. Using fuzzy PID control, the system dynamically adjusts motor speed, achieving row spacing consistency below 3.9 % and seeding stability within 1.3 %, even at varying speeds. This system addresses critical challenges in precision agriculture, enhancing planting efficiency and reducing labor costs. This innovation enhances planting efficiency, reduces labor costs, and ensures adaptability to varying tractor speeds, meeting the precision requirements of wheat planting.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 1","pages":"Pages 12-25"},"PeriodicalIF":8.2,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097979","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-22DOI: 10.1016/j.aiia.2024.12.002
Jie Guo , Zhou Yang , Manoj Karkee , Jieli Duan , Yong He
Banana de-handing is an important part of banana post-harvesting operation. The traditional artificial de-handing model has problems such as labor intensity, inaccurate cutting, uneven cutting surface, unstable catching, and damage of banana fruit, etc. The mapping relationship between the characteristic parameters of the movement posture of the cutter and the influencing factors of the contact stress of banana crown cutting in unstructured environments, and the changing rules of the bumping contact stress of complex multi-shaped banana fruit with the physical property parameters of the cushioning material are the theoretical and technical difficulties that urgently need to be solved in the realization of banana mechanical de-handing. The future research (research on the coupling mechanism of visual cognition-mechanism cutting and low-destructive catching method of full-field continuous de-handing of bananas under multi-constraint scenarios) should: (1) create a database of banana crown, obtain the optimal banana crown recognition model, develop a recognition and locating system of the cutting line of banana crown and obtain its spatial location information; (2) establish the discrete element mechanical model of banana crown and the interaction model between banana crown and the cutter, clarify the stress change and the force wave transmission characteristics of the cutting process, construct the multi-objective optimization equation of the cutting performance, obtain the best combination of cutting parameters, and ascertain the mechanisms of synergistic locating and continuous cutting of banana crown; (3) establish the contact mechanical model of banana fruit drop-bump, parse the bumping characteristics between banana fruit and cushioning material, construct mathematical equations to quantitatively assess damage results, and determine the detract catching method of banana fruit that matches the de-handing mode in multi-constraint scenarios. This study showed that the real-time identification and spatial positioning of fruit, the mechanical properties of crown and the optimization of cutting performance, the damage mechanism of fruit and its loss-reducing harvesting method are the three key breakthroughs in realizing the robotization of de-handing. The current bottleneck problems and future research directions of intelligent banana de-handing were pointed out in this paper, which can provide a theoretical basis for the optimal design of banana de-handing devices and provide technical support for promoting the practical application of intelligent de-handing equipment.
{"title":"Robotization of banana de-handing under multi-constraint scenarios: Challenges and future directions","authors":"Jie Guo , Zhou Yang , Manoj Karkee , Jieli Duan , Yong He","doi":"10.1016/j.aiia.2024.12.002","DOIUrl":"10.1016/j.aiia.2024.12.002","url":null,"abstract":"<div><div>Banana de-handing is an important part of banana post-harvesting operation. The traditional artificial de-handing model has problems such as labor intensity, inaccurate cutting, uneven cutting surface, unstable catching, and damage of banana fruit, etc. The mapping relationship between the characteristic parameters of the movement posture of the cutter and the influencing factors of the contact stress of banana crown cutting in unstructured environments, and the changing rules of the bumping contact stress of complex multi-shaped banana fruit with the physical property parameters of the cushioning material are the theoretical and technical difficulties that urgently need to be solved in the realization of banana mechanical de-handing. The future research (research on the coupling mechanism of visual cognition-mechanism cutting and low-destructive catching method of full-field continuous de-handing of bananas under multi-constraint scenarios) should: (1) create a database of banana crown, obtain the optimal banana crown recognition model, develop a recognition and locating system of the cutting line of banana crown and obtain its spatial location information; (2) establish the discrete element mechanical model of banana crown and the interaction model between banana crown and the cutter, clarify the stress change and the force wave transmission characteristics of the cutting process, construct the multi-objective optimization equation of the cutting performance, obtain the best combination of cutting parameters, and ascertain the mechanisms of synergistic locating and continuous cutting of banana crown; (3) establish the contact mechanical model of banana fruit drop-bump, parse the bumping characteristics between banana fruit and cushioning material, construct mathematical equations to quantitatively assess damage results, and determine the detract catching method of banana fruit that matches the de-handing mode in multi-constraint scenarios. This study showed that the real-time identification and spatial positioning of fruit, the mechanical properties of crown and the optimization of cutting performance, the damage mechanism of fruit and its loss-reducing harvesting method are the three key breakthroughs in realizing the robotization of de-handing. The current bottleneck problems and future research directions of intelligent banana de-handing were pointed out in this paper, which can provide a theoretical basis for the optimal design of banana de-handing devices and provide technical support for promoting the practical application of intelligent de-handing equipment.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 1","pages":"Pages 1-11"},"PeriodicalIF":8.2,"publicationDate":"2024-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143097983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-19DOI: 10.1016/j.aiia.2024.12.003
Chen Zhenyu , Dou Hanjie , Gao Yuanyuan , Zhai Changyuan , Wang Xiu , Zou Wei
Orchard intelligent equipment must perform autonomous navigation tasks along fruit tree row centrelines and headlands according to established operational requirements. The tree canopy obstructs satellite signals, limiting the accuracy and stability of the GNSS-based autonomous navigation system. This paper presents a multipoint autonomous navigation method with the orchard row centreline navigation capabilities by integrating light detection and ranging (LiDAR) and inertial measurement unit (IMU) data. The method begins by constructing a three-dimensional (3D) point cloud map of the orchard via the LIO_SAM algorithm, and a 3D point cloud-to-two-dimensional (2D) grid map algorithm is designed. This algorithm retains the tree trunk position information from the point cloud based on tree trunk features to obtain a 2D grid map for orchard navigation, and the navigation point coordinates were calculated based on tree trunk positions. A multipoint navigation method was designed, where the system automatically determines the completion status of the previous navigation point and sequentially issues navigation point coordinates, enabling autonomous navigation along the row centrelines and headlands during orchard operations. Row centreline navigation tests and headland turning tests were conducted, and the performances of 16-line and 32-line LiDAR with this method are compared. The research results reveal that the multipoint navigation method could achieve movement along orchard row centrelines and deploy autonomous turning. The 32-line LiDAR data demonstrated an average absolute lateral deviation of 1.83 cm, a standard deviation of 1.60 cm, and a maximum deviation of 10.30 cm at a 3-m navigation point interval, indicating greater precision. However, the turning time was longer, with increases of 8.11 % and 6.13 % with the two different turning methods compared to the 16-line LiDAR. The research results provide support for research on autonomous navigation technology for intelligent orchard equipment.
{"title":"Research on an orchard row centreline multipoint autonomous navigation method based on LiDAR","authors":"Chen Zhenyu , Dou Hanjie , Gao Yuanyuan , Zhai Changyuan , Wang Xiu , Zou Wei","doi":"10.1016/j.aiia.2024.12.003","DOIUrl":"10.1016/j.aiia.2024.12.003","url":null,"abstract":"<div><div>Orchard intelligent equipment must perform autonomous navigation tasks along fruit tree row centrelines and headlands according to established operational requirements. The tree canopy obstructs satellite signals, limiting the accuracy and stability of the GNSS-based autonomous navigation system. This paper presents a multipoint autonomous navigation method with the orchard row centreline navigation capabilities by integrating light detection and ranging (LiDAR) and inertial measurement unit (IMU) data. The method begins by constructing a three-dimensional (3D) point cloud map of the orchard via the LIO_SAM algorithm, and a 3D point cloud-to-two-dimensional (2D) grid map algorithm is designed. This algorithm retains the tree trunk position information from the point cloud based on tree trunk features to obtain a 2D grid map for orchard navigation, and the navigation point coordinates were calculated based on tree trunk positions. A multipoint navigation method was designed, where the system automatically determines the completion status of the previous navigation point and sequentially issues navigation point coordinates, enabling autonomous navigation along the row centrelines and headlands during orchard operations. Row centreline navigation tests and headland turning tests were conducted, and the performances of 16-line and 32-line LiDAR with this method are compared. The research results reveal that the multipoint navigation method could achieve movement along orchard row centrelines and deploy autonomous turning. The 32-line LiDAR data demonstrated an average absolute lateral deviation of 1.83 cm, a standard deviation of 1.60 cm, and a maximum deviation of 10.30 cm at a 3-m navigation point interval, indicating greater precision. However, the turning time was longer, with increases of 8.11 % and 6.13 % with the two different turning methods compared to the 16-line LiDAR. The research results provide support for research on autonomous navigation technology for intelligent orchard equipment.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"15 2","pages":"Pages 221-231"},"PeriodicalIF":8.2,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate classification of cassava disease, particularly in field scenarios, relies on object semantic localization to identify and precisely locate specific objects within an image based on their semantic meaning, thereby enabling targeted classification while suppressing irrelevant noise and focusing on key semantic features. The advancement of deep convolutional neural networks (CNNs) paved the way for identifying cassava diseases by leveraging salient semantic features and promising high returns. This study proposes an approach that incorporates three innovative elements to refine feature representation for cassava disease classification. First, a mutualattention method is introduced to highlight semantic features and suppress irrelevant background features in the feature maps. Second, instance batch normalization (IBN) was employed after the residual unit to construct salient semantic features using the mutualattention method, representing high-quality semantic features in the foreground. Finally, the RSigELUD activation method replaced the conventional ReLU activation, enhancing the nonlinear mapping capacity of the proposed neural network and further improving fine-grained leaf disease classification performance. This approach significantly aided in distinguishing subtle disease manifestations in cassava leaves. The proposed neural network, MAIRNet-101 (Mutualattention IBN RSigELUD Neural Network), achieved an accuracy of 95.30 % and an F1-score of 0.9531, outperforming EfficientNet-B5 and RepVGG-B3g4. To evaluate the generalization capability of MAIRNet, the FGVC-Aircraft dataset was used to train MAIRNet-50, which achieved an accuracy of 83.64 %. These results suggest that the proposed algorithm is well suited for cassava leaf disease classification applications and offers a robust solution for advancing agricultural technology.
{"title":"A salient feature establishment tactic for cassava disease recognition","authors":"Jiayu Zhang , Baohua Zhang , Zixuan Chen , Innocent Nyalala , Kunjie Chen , Junfeng Gao","doi":"10.1016/j.aiia.2024.11.004","DOIUrl":"10.1016/j.aiia.2024.11.004","url":null,"abstract":"<div><div>Accurate classification of cassava disease, particularly in field scenarios, relies on object semantic localization to identify and precisely locate specific objects within an image based on their semantic meaning, thereby enabling targeted classification while suppressing irrelevant noise and focusing on key semantic features. The advancement of deep convolutional neural networks (CNNs) paved the way for identifying cassava diseases by leveraging salient semantic features and promising high returns. This study proposes an approach that incorporates three innovative elements to refine feature representation for cassava disease classification. First, a mutualattention method is introduced to highlight semantic features and suppress irrelevant background features in the feature maps. Second, instance batch normalization (IBN) was employed after the residual unit to construct salient semantic features using the mutualattention method, representing high-quality semantic features in the foreground. Finally, the RSigELUD activation method replaced the conventional ReLU activation, enhancing the nonlinear mapping capacity of the proposed neural network and further improving fine-grained leaf disease classification performance. This approach significantly aided in distinguishing subtle disease manifestations in cassava leaves. The proposed neural network, MAIRNet-101 (Mutualattention IBN RSigELUD Neural Network), achieved an accuracy of 95.30 % and an F1-score of 0.9531, outperforming EfficientNet-B5 and RepVGG-B3g4. To evaluate the generalization capability of MAIRNet, the FGVC-Aircraft dataset was used to train MAIRNet-50, which achieved an accuracy of 83.64 %. These results suggest that the proposed algorithm is well suited for cassava leaf disease classification applications and offers a robust solution for advancing agricultural technology.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"14 ","pages":"Pages 115-132"},"PeriodicalIF":8.2,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143153819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.aiia.2024.11.005
Mohammad Amin Razavi , A. Pouyan Nejadhashemi , Babak Majidi , Hoda S. Razavi , Josué Kpodo , Rasu Eeswaran , Ignacio Ciampitti , P.V. Vara Prasad
In this study, we employ advanced data-driven techniques to investigate the complex relationships between the yields of five major crops and various geographical and spatiotemporal features in Senegal. We analyze how these features influence crop yields by utilizing remotely sensed data. Our methodology incorporates clustering algorithms and correlation matrix analysis to identify significant patterns and dependencies, offering a comprehensive understanding of the factors affecting agricultural productivity in Senegal. To optimize the model's performance and identify the optimal hyperparameters, we implemented a comprehensive grid search across four distinct machine learning regressors: Random Forest, Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), and Light Gradient-Boosting Machine (LightGBM). Each regressor offers unique functionalities, enhancing our exploration of potential model configurations. The top-performing models were selected based on evaluating multiple performance metrics, ensuring robust and accurate predictive capabilities. The results demonstrated that XGBoost and CatBoost perform better than the other two. We introduce synthetic crop data generated using a Variational Auto Encoder to address the challenges posed by limited agricultural datasets. By achieving high similarity scores with real-world data, our synthetic samples enhance model robustness, mitigate overfitting, and provide a viable solution for small dataset issues in agriculture. Our approach distinguishes itself by creating a flexible model applicable to various crops together. By integrating five crop datasets and generating high-quality synthetic data, we improve model performance, reduce overfitting, and enhance realism. Our findings provide crucial insights for productivity drivers in key cropping systems, enabling robust recommendations and strengthening the decision-making capabilities of policymakers and farmers in data-scarce regions.
{"title":"Enhancing crop yield prediction in Senegal using advanced machine learning techniques and synthetic data","authors":"Mohammad Amin Razavi , A. Pouyan Nejadhashemi , Babak Majidi , Hoda S. Razavi , Josué Kpodo , Rasu Eeswaran , Ignacio Ciampitti , P.V. Vara Prasad","doi":"10.1016/j.aiia.2024.11.005","DOIUrl":"10.1016/j.aiia.2024.11.005","url":null,"abstract":"<div><div>In this study, we employ advanced data-driven techniques to investigate the complex relationships between the yields of five major crops and various geographical and spatiotemporal features in Senegal. We analyze how these features influence crop yields by utilizing remotely sensed data. Our methodology incorporates clustering algorithms and correlation matrix analysis to identify significant patterns and dependencies, offering a comprehensive understanding of the factors affecting agricultural productivity in Senegal. To optimize the model's performance and identify the optimal hyperparameters, we implemented a comprehensive grid search across four distinct machine learning regressors: Random Forest, Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), and Light Gradient-Boosting Machine (LightGBM). Each regressor offers unique functionalities, enhancing our exploration of potential model configurations. The top-performing models were selected based on evaluating multiple performance metrics, ensuring robust and accurate predictive capabilities. The results demonstrated that XGBoost and CatBoost perform better than the other two. We introduce synthetic crop data generated using a Variational Auto Encoder to address the challenges posed by limited agricultural datasets. By achieving high similarity scores with real-world data, our synthetic samples enhance model robustness, mitigate overfitting, and provide a viable solution for small dataset issues in agriculture. Our approach distinguishes itself by creating a flexible model applicable to various crops together. By integrating five crop datasets and generating high-quality synthetic data, we improve model performance, reduce overfitting, and enhance realism. Our findings provide crucial insights for productivity drivers in key cropping systems, enabling robust recommendations and strengthening the decision-making capabilities of policymakers and farmers in data-scarce regions.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"14 ","pages":"Pages 99-114"},"PeriodicalIF":8.2,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142742953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Due to intensive genetic selection for rapid growth rates and high broiler yields in recent years, the global poultry industry has faced a challenging problem in the form of woody breast (WB) conditions. This condition has caused significant economic losses as high as $200 million annually, and the root cause of WB has yet to be identified. Human palpation is the most common method of distinguishing a WB from others. However, this method is time-consuming and subjective. Hyperspectral imaging (HSI) combined with machine learning algorithms can evaluate the WB conditions of fillets in a non-invasive, objective, and high-throughput manner. In this study, 250 raw chicken breast fillet samples (normal, mild, severe) were taken, and spatially heterogeneous hardness distribution was first considered when designing HSI processing models. The study not only classified the WB levels from HSI but also built a regression model to correlate the spectral information with sample hardness data. To achieve a satisfactory classification and regression model, a neural network architecture search (NAS) enabled a wide-deep neural network model named NAS-WD, which was developed. In NAS-WD, NAS was first used to automatically optimize the network architecture and hyperparameters. The classification results show that NAS-WD can classify the three WB levels with an overall accuracy of 95 %, outperforming the traditional machine learning model, and the regression correlation between the spectral data and hardness was 0.75, which performs significantly better than traditional regression models.
{"title":"Neural network architecture search enabled wide-deep learning (NAS-WD) for spatially heterogenous property awared chicken woody breast classification and hardness regression","authors":"Chaitanya Pallerla , Yihong Feng , Casey M. Owens , Ramesh Bahadur Bist , Siavash Mahmoudi , Pouya Sohrabipour , Amirreza Davar , Dongyi Wang","doi":"10.1016/j.aiia.2024.11.003","DOIUrl":"10.1016/j.aiia.2024.11.003","url":null,"abstract":"<div><div>Due to intensive genetic selection for rapid growth rates and high broiler yields in recent years, the global poultry industry has faced a challenging problem in the form of woody breast (WB) conditions. This condition has caused significant economic losses as high as $200 million annually, and the root cause of WB has yet to be identified. Human palpation is the most common method of distinguishing a WB from others. However, this method is time-consuming and subjective. Hyperspectral imaging (HSI) combined with machine learning algorithms can evaluate the WB conditions of fillets in a non-invasive, objective, and high-throughput manner. In this study, 250 raw chicken breast fillet samples (normal, mild, severe) were taken, and spatially heterogeneous hardness distribution was first considered when designing HSI processing models. The study not only classified the WB levels from HSI but also built a regression model to correlate the spectral information with sample hardness data. To achieve a satisfactory classification and regression model, a neural network architecture search (NAS) enabled a wide-deep neural network model named NAS-WD, which was developed. In NAS-WD, NAS was first used to automatically optimize the network architecture and hyperparameters. The classification results show that NAS-WD can classify the three WB levels with an overall accuracy of 95 %, outperforming the traditional machine learning model, and the regression correlation between the spectral data and hardness was 0.75, which performs significantly better than traditional regression models.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"14 ","pages":"Pages 73-85"},"PeriodicalIF":8.2,"publicationDate":"2024-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142699967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-14DOI: 10.1016/j.aiia.2024.11.001
Fatima K. Abu Salem , Sara Awad , Yasmine Hamdar , Samer Kharroubi , Hadi Jaafar
Estimating actual evapotranspiration (ETₐ) is crucial for water resource management, yet existing methods face limitations. Traditional approaches, including eddy covariance and remote sensing-based energy balance methods, often struggle with high costs, limited spatial and temporal coverage, and reduced predictive accuracy, particularly for classical empirical models. While machine learning has emerged as a promising alternative, it still presents challenges, notably in underestimating ETₐ during periods of high heat. We attribute this to insufficient learning on the rare but highly relevant ETₐ values of interest, or the not-so-big climatic datasets available for use. In this manuscript, we demonstrate how few-shot, meta-learning models (MAML) that are specifically designed for enhanced generalizability on not-so-big datasets can outperform basic machine learning models in upscaling ETₐ from two major in-situ towers, the Ameriflux and Euroflux. Using limited remotely sensed land surface data from the METRIC-EEFlux and limited climatic variables, we demonstrate that the chosen models can attain quantifiable utility within the utility-based-regression paradigm towards impactful practical considerations. Our initial explorations reveal that EEflux ETₐ deviates significantly from in-situ observations measured through the Ameriflux and EEflux towers (). Instead, MAML shows best performance in approximating ETₐ than basic machine learning algorithms and EEFlux ( on entire testing dataset, on the Csa climate, on the Cfa climate, and on the CSH vegetation class), and continues to improve without overfitting even when exposed to a relatively small training dataset. Its high F2 score (96 %) indicates that MAML has very high precision and recall for rare cases, which is significant for irrigation. Of independent interest, this study confirms that limited remotely sensed EEflux products contribute significantly to knowledge about ground truth ETₐ and can thus be of valuable use in settings where access to good quality and high-volume data is compromised.
{"title":"Utility-based regression and meta-learning techniques for modeling actual ET: Comparison to (METRIC-EEFLUX) model","authors":"Fatima K. Abu Salem , Sara Awad , Yasmine Hamdar , Samer Kharroubi , Hadi Jaafar","doi":"10.1016/j.aiia.2024.11.001","DOIUrl":"10.1016/j.aiia.2024.11.001","url":null,"abstract":"<div><div>Estimating actual evapotranspiration (ETₐ) is crucial for water resource management, yet existing methods face limitations. Traditional approaches, including eddy covariance and remote sensing-based energy balance methods, often struggle with high costs, limited spatial and temporal coverage, and reduced predictive accuracy, particularly for classical empirical models. While machine learning has emerged as a promising alternative, it still presents challenges, notably in underestimating ETₐ during periods of high heat. We attribute this to insufficient learning on the rare but highly relevant ETₐ values of interest, or the not-so-big climatic datasets available for use. In this manuscript, we demonstrate how <em>few-shot, meta-learning models (MAML)</em> that are specifically designed for enhanced generalizability on not-so-big datasets can outperform basic machine learning models in upscaling ETₐ from two major in-situ towers, the Ameriflux and Euroflux. Using limited remotely sensed land surface data from the METRIC-EEFlux and limited climatic variables, we demonstrate that the chosen models can attain quantifiable utility within the <em>utility-based-regression</em> paradigm towards impactful practical considerations. Our initial explorations reveal that EEflux ETₐ deviates significantly from in-situ observations measured through the Ameriflux and EEflux towers (<span><math><msup><mi>R</mi><mn>2</mn></msup><mo>=</mo><mn>39</mn><mo>%</mo></math></span>). Instead, MAML shows best performance in approximating ETₐ than basic machine learning algorithms and EEFlux (<span><math><msup><mi>R</mi><mn>2</mn></msup><mo>=</mo><mn>71</mn><mo>%</mo></math></span> on entire testing dataset, <span><math><msup><mi>R</mi><mn>2</mn></msup><mo>=</mo><mn>0.88</mn></math></span> on the Csa climate, <span><math><msup><mi>R</mi><mn>2</mn></msup><mo>=</mo><mn>0.79</mn></math></span> on the Cfa climate, and <span><math><msup><mi>R</mi><mn>2</mn></msup><mo>=</mo><mn>0.78</mn></math></span> on the CSH vegetation class), and continues to improve without overfitting even when exposed to a relatively small training dataset. Its high F2 score (96 %) indicates that MAML has very high precision and recall for rare cases, which is significant for irrigation. Of independent interest, this study confirms that limited remotely sensed EEflux products contribute significantly to knowledge about ground truth ETₐ and can thus be of valuable use in settings where access to good quality and high-volume data is compromised.</div></div>","PeriodicalId":52814,"journal":{"name":"Artificial Intelligence in Agriculture","volume":"14 ","pages":"Pages 43-55"},"PeriodicalIF":8.2,"publicationDate":"2024-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142699966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号