Computers and Electronics in Agriculture最新文献

{"title":"Analyzing maize stem circumference, stem height, and stem circumference-to-height ratio using UAV, UGV, and deep learning","authors":"Jianing Shen ,&nbsp;Jibo Yue ,&nbsp;Yang Liu ,&nbsp;Yihan Yao ,&nbsp;Haikuan Feng ,&nbsp;Hao Yang ,&nbsp;Wei Guo ,&nbsp;Xinming Ma ,&nbsp;Yuanyuan Fu ,&nbsp;Meiyan Shu ,&nbsp;Guijun Yang ,&nbsp;Hongbo Qiao","doi":"10.1016/j.compag.2025.111019","DOIUrl":"10.1016/j.compag.2025.111019","url":null,"abstract":"<div><div>Maize (Zea mays L.) is a crucial grain and economic crop with extensive applications in food, feed, and industry. Phenotypic traits such as stem circumference (SC), stem height (SH), and the stem circumference-to-height ratio (SCHR) are essential indicators for studying maize development, environmental adaptation, and lodging resistance. Traditional manual measurement methods are inefficient, costly, and unsuitable for large-scale phenotypic monitoring. While unmanned aerial vehicle (UAV)-based approaches have achieved relatively accurate SH estimation, SC estimation remains challenging using UAV technology alone, limiting SCHR estimation. This study combined digital camera sensors on unmanned ground vehicle (UGV) and UAV platforms to capture maize stem and canopy images, enabling the estimation of SC, SH, and SCHR. The primary contributions of this study are as follows: (1) We propose the maize-stem segmentation network for calculating stem diameter and circumference (MSSDCNet) to segment maize stems in images and estimate SC based on the segmentation results. (2) We process UAV-derived digital surface models to extract SH information and employ a linear regression (LR) model for SH estimation. (3) Using the estimated SC and SH, we calculate SCHR and analyze its temporal variations across different growth stages. The results demonstrate that: (1) MSSDCNet accurately segments maize stems from images and facilitates SC estimation (<em>R²</em> = 0.759, <em>RMSE</em> = 0.414 cm, <em>nRMSE</em> = 0.087). Temporal analysis of SC reveals a gradual decrease during the reproductive growth stage, potentially due to the transfer of photosynthetic products to the maize cob and stem water loss. (2) This study accurately estimated SH (<em>R</em>² = 0.941, <em>RMSE</em> = 0.151 m, <em>nRMSE</em> = 0.078). However, SH estimates during the reproductive growth stage tend to be underestimated, likely due to DSM point clouds being more sensitive to sharp features such as tassels. (3) SCHR estimation achieves <em>R</em>² = 0.453, <em>RMSE</em> = 2.287 × 10⁻³, <em>nRMSE</em> = 0.136. Temporal analysis reveals a general decline in SCHR from the kernel blister stage to the dough stage, with some maize materials consistently exhibiting lower SCHR levels during each growth stage. This may be related to genetic traits, planting density, soil fertility, or other environmental factors. By integrating UGV-based maize stem images and UAV-based maize canopy images with MSSDCNet and LR, this study successfully estimates SC, SH, and SCHR for various maize materials. This study provides a novel technique for maize, early lodging risk prediction, and lodging-resistant breeding lines screening, contributing to rapid and efficient maize phenotypic monitoring under field conditions.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111019"},"PeriodicalIF":8.9,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145118620","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}

{"title":"CT-HiffNet: A contour-texture hierarchical feature fusion network for cropland field parcel extraction from high-resolution remote sensing images","authors":"Hao Wu ,&nbsp;Junyang Xie ,&nbsp;Weihao Deng ,&nbsp;Anqi Lin ,&nbsp;Abdul Rashid Mohamed Shariff ,&nbsp;Shamshodbek Akmalov ,&nbsp;Wenbin Wu ,&nbsp;Zhaoliang Li ,&nbsp;Qiangyi Yu ,&nbsp;Qunming Wang ,&nbsp;Jian Zhang ,&nbsp;Xin Mei ,&nbsp;Qiong Hu","doi":"10.1016/j.compag.2025.111010","DOIUrl":"10.1016/j.compag.2025.111010","url":null,"abstract":"<div><div>Automatically extracting cropland field parcels from remote sensing images is crucial for developing smart agriculture. However, notable spatio-spectral differences captured by multiple remote sensing sensors at different times led to the uncertain contour and texture features among large-scale cropland field parcel, posing challenges for robust and high-precision extraction. To address these challenges, we proposed a contour-texture hierarchical feature fusion network (CT-HiffNet) for cropland field parcels extraction from high-resolution remote sensing images. The CT-HiffNet consists of three modules: a hybrid module integrating attention and guidance method to thoroughly learn the internal texture features as well as external contour features of cropland field parcels; a deep residual shrinkage block for feature encoding to effectively eliminate redundant information during the extraction tasks; and a hierarchical information fusion decoder to enhance contour-texture feature interactions at different scales and minimize information loss during feature restoration. The CT-HiffNet was evaluated across four distinct agricultural landscape regions in China using GaoFen-2 images, as well as in six other global regions using Sentinel-2 and Google Earth images. The results show that CT-HiffNet achieves OA, precision, and recall all exceeding 80% across various regions in China, and in other global validation areas, precision and recall surpass 84% and 86.5%, respectively. This demonstrates its effectiveness in extracting cropland field parcels and indicates the model’s strong transferability and generalization capability. In particularly, the contour–texture feature effectively enhanced the boundary recognition of cropland field parcels, contributing to the model adaptability to different acquirement times of remote sensing images. Meanwhile, determining an appropriate sample size is crucial for the performance of CT-HiffNet.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111010"},"PeriodicalIF":8.9,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145118686","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}

{"title":"Novel soil health assessment framework for legume-based rotation farmland by interpretable machine learning with causal inference","authors":"Xuebin Xu ,&nbsp;Qiong Liu ,&nbsp;Yalin Liu ,&nbsp;Yongfu Li ,&nbsp;Yixuan Chen ,&nbsp;Tong Lei ,&nbsp;Yakov Kuzyakov ,&nbsp;Wenju Zhang ,&nbsp;Jianping Chen ,&nbsp;Tida Ge","doi":"10.1016/j.compag.2025.111011","DOIUrl":"10.1016/j.compag.2025.111011","url":null,"abstract":"<div><div>Accurate and robust soil health assessment is essential for sustaining legume-based rotation systems and informing their optimized management. To address the limitations of conventional methods in capturing management-induced variations, we developed an innovative framework grounded in the theoretical hypothesis that soil health reflects soil’s capacity to maximize production stability while minimizing input requirements. This framework synergistically integrates interpretable machine learning with causal inference and network analysis (CI-SHAP-NA), implementing a systematic workflow encompassing indicator selection, quantitative scoring, and multidimensional integration. Our framework was systematically implemented to assess soil health across diverse legume-based rotation systems in China. The results showed that CI-SHAP-NA identified a parsimonious yet highly informative set of indicators (soil organic carbon, available iron, and cellobiohydrolase) demonstrating superior explanatory power for critical soil ecological processes. The derived soil health index (SHI) by the CI-SHAP-NA framework demonstrated enhanced discriminative capacity (SHI range: 0.01−0.92) and strong concordance (R² = 0.80) with conventional total dataset assessment while maintaining significant predictive validity for crop productivity (Pearson’s <em>r</em> = 0.21, <em>p</em> &lt; 0.001). It consistently outperformed PCA and NA methods in both explanatory power and fairness comparisons. The selected indicators proved robust and non-redundant, as substituting any indicator significantly reduced the correlation and sensitivity of SHI. Furthermore, CI-SHAP-NA demonstrated strong transferability, showing a stronger correlation with yield (<em>r</em> = 0.25, <em>p</em> &lt; 0.001) on internally established independent sites than PCA and NA. This framework successfully resolved previously obscured soil health gradients between contrasting management systems, with paddy-legume rotations consistently outperforming their dryland counterparts − a differentiation rigorously validated against traditional benchmarks. These findings collectively establish the CI-SHAP-NA framework as a transformative tool for soil health assessment, offering substantial advantages over conventional approaches in terms of analytical robustness, ecological relevance, and practical utility. Future research should aim to incorporate multi-functional indicators as well as evaluate the framework’s performance across varied agroecosystems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111011"},"PeriodicalIF":8.9,"publicationDate":"2025-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105859","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}

{"title":"A 3D phenotyping pipeline for peanut plants using point cloud","authors":"Bo Zhang ,&nbsp;Xin Yang ,&nbsp;Xiaodan Han ,&nbsp;Guowei Li ,&nbsp;Xianju Lu ,&nbsp;Bo Bai ,&nbsp;Haisheng Liu ,&nbsp;Teng Miao ,&nbsp;Sheng Wu ,&nbsp;Xinyu Guo","doi":"10.1016/j.compag.2025.110986","DOIUrl":"10.1016/j.compag.2025.110986","url":null,"abstract":"<div><div>Three-dimensional phenotyping technology is paramount in the field of peanut breeding and cultivation. The intricate topological structure of plants substantially complicates the development of effective peanut phenotyping technologies. In this study, we present the development of a point-cloud-based pipeline for three-dimensional phenotypic analysis of peanut plants. An efficient multi-view image acquisition system and three-dimensional reconstruction techniques were employed to generate point clouds of peanut plants. A dataset comprising 188 labelled samples of peanut point clouds was constructed for the development of semantic and leaf-instance segmentation models based on the transformer architecture. The segmentation accuracy of these models surpassed that of the conventional general segmentation techniques for plant point clouds. Based on the results of the segmentation, 11 three-dimensional phenotypic traits were automatically calculated at both the plant and leaf scales. Among these, five phenotypic traits, including plant height and leaf length, exhibited a mean absolute percentage error (MAPE) of less than 0.12 compared to the measured values. In addition, the Jensen-Shannon divergence (JS divergence) between the probability distributions of the three leaf phenotypic traits and their corresponding measured values was below 0.1. The three-dimensional phenotypic analysis pipeline developed in this study exhibited satisfactory generalisation capabilities, thereby offering an efficacious and expeditious high-throughput phenotyping analysis instrument for the intelligent breeding and cultivation of peanuts.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110986"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105504","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}

{"title":"Real-time ion concentration pattern analysis in plants based on microneedle-type sensing and time-series prediction","authors":"Jiawei Zhai ,&nbsp;Bin Luo ,&nbsp;Hongtu Dong ,&nbsp;Aixue Li ,&nbsp;Xiaotong Jin ,&nbsp;Chunjiang Zhao ,&nbsp;Xiaodong Wang","doi":"10.1016/j.compag.2025.111012","DOIUrl":"10.1016/j.compag.2025.111012","url":null,"abstract":"<div><div>In situ detection of plant ion signals faces technical limitations in terms of real-time capability, minimal invasiveness, and data analysis. Therefore, the development of sensors for in vivo plant detection and construction of time-series prediction models to analyze the dynamic patterns of ion concentrations in plants are imperative. This study presents a microneedle electrode system for potassium ion (K⁺) sensing, which is applied to real-time in situ detection in lettuce. The microneedle ion-selective electrodes (ISEs) fabricated herein exhibited a rapid potentiometric response (within &lt; 15 s), with concentration responses adhering to the Nernst equation. During in vivo plant detection, the system captured instantaneous ion-signal changes upon exogenous application without influencing subsequent plant growth. This study demonstrates the pioneering application of time-series prediction (nonlinear autoregressive neural network model) to analyze in vivo K⁺ signals in lettuce, accurately forecasting ion concentration dynamics over time and identifying the transition pattern from signal fluctuation to stabilization. The integration of microneedle ISE-based in situ plant monitoring with time-series prediction represents a crucial and reliable approach to agricultural sensor innovation, providing a novel paradigm for precision agriculture and plant stress response research.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111012"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105497","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}

{"title":"From rolling-in to enveloping: An active gripper for efficient handling of spherical fruits in agriculture","authors":"Huanhuan Qin ,&nbsp;Zicheng Qiu ,&nbsp;Haoran Zhu ,&nbsp;Rui Diao ,&nbsp;Fengwei Che ,&nbsp;Xingjian Gu ,&nbsp;Yan Luo ,&nbsp;Baohua Zhang ,&nbsp;Mingzhou Lu","doi":"10.1016/j.compag.2025.110995","DOIUrl":"10.1016/j.compag.2025.110995","url":null,"abstract":"<div><div>Robotic grippers enable robots to perform agricultural product handling tasks, such as grasping, holding, and lifting, reducing the reliance on human labor. Currently, most agricultural grippers adopt force or form closure modes, which are sensitive to the position of product. In this paper, an active rolling gripper is proposed for spherical fruit handling. It utilizes an innovative rolling-in and enveloping mode, with advantages in terms of high efficiency and high position offset tolerance. The gripper is equipped with three fingers and an elastic membrane, with active rollers at the fingertips. The active roller allows the fruit to be guided into the gripper with only a gentle touch. The elastic membrane, together with the fingers, provides support and restraint when the fruit is fully inside the gripper. To demonstrate the working principle of the gripper, the gripper-fruit interaction model was derived. Then, three experiments were conducted to evaluate its actual fruit handling capability. The results proved that the proposed gripper had high vertical and horizontal offset tolerance when handling tomatoes, oranges, and apples with different orientations. The gripper could complete the rolling-in and enveloping process within 0.5 s once it contacted the fruits. The effectiveness of the gripper was also verified in real-field citrus picking scenarios.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110995"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105503","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}

{"title":"Exploring fruit mechanical injury mechanisms: a review of numerical simulation techniques in the whole supply chain","authors":"Shujie Gao ,&nbsp;Xiaowei Huang ,&nbsp;Zhihua Li ,&nbsp;Xinai Zhang ,&nbsp;Zhecong Yuan ,&nbsp;Hany S. El-Mesery ,&nbsp;Jiyong Shi ,&nbsp;Xiaobo Zou","doi":"10.1016/j.compag.2025.110998","DOIUrl":"10.1016/j.compag.2025.110998","url":null,"abstract":"<div><div>Mechanical injury is a key factor causing postharvest quality deterioration of fruits, significantly affecting their storage life and market competitiveness. Such injuries can occur during harvesting, packaging, processing, and transportation. A thorough understanding of the factors influencing fresh fruit damage is crucial for developing effective postharvest loss reduction strategies. Numerical simulation technologies, especially the finite element method (FEM), extended finite element method (XFEM), and discrete element method (DEM), provide powerful analytical tools for studying the response of fruits to compression, vibration, and impact under various mechanical loads and environmental conditions. Although these methods cannot directly prevent mechanical damage, they can offer scientific evidence and technical support for optimizing packaging structures, transportation methods, and harvesting mechanisms, thereby minimizing potential risks. This review systematically examines the formation mechanisms and physiological consequences of mechanical injuries in fruits, discusses the principles and latest application progress of mainstream numerical simulation methods, and focuses on key aspects of finite element analysis, such as model construction, material parameter setting, mesh generation, and boundary condition definition. It also explores the potential applications of simulation-driven design in various stages of the fruit supply chain. In the future, combining the physiological characteristics of fruits with artificial intelligence modeling may enable real-time decision-making and intelligent regulation of postharvest systems, promoting the development of fruit circulation toward high quality, low loss, and sustainability. By integrating fruit biomechanics and digital modeling, numerical simulations provide a solid theoretical foundation for minimizing fruit quality and value losses.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110998"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105862","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}

{"title":"Tightly coupled GNSS/IMU/vision integrated system for positioning in agricultural scenarios","authors":"Jiayuan Yu ,&nbsp;Hui Fang ,&nbsp;Xiya Zhang ,&nbsp;Wentao Wu ,&nbsp;Yong He","doi":"10.1016/j.compag.2025.110478","DOIUrl":"10.1016/j.compag.2025.110478","url":null,"abstract":"<div><div>The traditional satellite-based positioning technology is widely used in the field of automatic navigation of agricultural machinery, but the meter-level accuracy of the Global Navigation Satellite System (GNSS) single-point positioning does not meet the needs of agriculture. Although the Real-Time Kinematic (RTK) carrier-phase differential positioning technology can achieve centimeter-level accuracy, it is associated with high costs. Compared with the GNSS-only positioning, multi-sensor fusion, which integrates cameras, Inertial Measurement Units (IMU), and other inexpensive sensors, offers significant advantages. This paper proposes a low-cost multi-sensor fusion system suitable for positioning in agricultural scenarios, which tightly couples information from GNSS, IMU, and vision. The raw data from each sensor are preprocessed, including feature tracking, IMU pre-integration, and screening of unhealthy, low-elevation satellites. A factor graph-based optimization model is developed to derive a drift-free global trajectory estimation. The factor graph incorporates visual, IMU, code pseudo-range, Doppler, and clock factors. To adapt to agricultural environments, two improvements to the visual factor are made. We introduce an IMU-assisted optical flow method to mitigate the impact of dynamic noise, such as wind-blown crops and pedestrians, on feature tracking. Additionally, we eliminate the inverse depth state quantity of the far-away feature points during the factor graph optimization process to reduce translation and scale errors introduced in the pose optimization. The system was tested in three agricultural scenarios, yielding results that demonstrated the ability of the approach to achieve an absolute localization accuracy of 0.87 m and a relative localization accuracy of 0.090 m.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110478"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105501","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}

{"title":"Lane navigation control method and equipment of chicken house based on 2D LiDAR","authors":"Yuxiao Han ,&nbsp;Yajun An ,&nbsp;Shuai Li ,&nbsp;Ning Wang ,&nbsp;Yuanyi Niu ,&nbsp;Man Zhang ,&nbsp;Han Li","doi":"10.1016/j.compag.2025.110993","DOIUrl":"10.1016/j.compag.2025.110993","url":null,"abstract":"<div><div>To enhance the efficiency of poultry farm management and reduce labor intensity, this study developed an autonomous inspection robot, named Poultry-Patrolman, for operation in high-density stacked-cage poultry houses. To address the challenges of precise navigation within narrow operation lanes, a comprehensive perception and control framework was proposed, with emphasis on data preprocessing, edge fitting, and adaptive control strategies. On the perception front, raw two-dimensional (2D) LiDAR data were transformed from polar to Cartesian coordinates and corrected for motion distortion based on odometry measurements between consecutive frames. For robust lane boundary extraction, a Full Sample Consensus (F-SAC) algorithm was proposed and applied to the segmented cloud points to perform edge fitting, from which a linear navigation line was generated to compute real-time deviation. On the control side, a Collaborative Hybrid Genetic-Particle Swarm Optimization (CHGAPSO) algorithm was employed to optimize the parameters of a PID controller. The optimized PID parameters, together with the navigation deviation, were integrated into an EKF-PID framework to achieve smooth and accurate trajectory tracking. Experimental results demonstrate that the F-SAC algorithm achieved a maximum absolute angular error of 2.328°, an average angular error of 0.116°, and a line fitting accuracy of 98.3 %. The CHGAPSO algorithm outperformed other methods in optimizing control parameters across four trajectory types: straight line, sinusoidal curve, composite curve, and noisy straight line. Furthermore, the EKF-PID control system demonstrated stable lane-following performance, consistently maintaining lateral steady-state errors within 2 cm under various initial poses at speeds of 0.2 m/s, 0.4 m/s, and 0.6 m/s. These findings validate the effectiveness and reliability of the proposed navigation framework for autonomous poultry house inspection.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110993"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105502","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}

{"title":"EMGCM: ensemble learning of multiple graph convolutional models for fish skeleton-based swimming behavior recognition","authors":"Beibei Li ,&nbsp;Mingrui Kong ,&nbsp;Yiran Liu ,&nbsp;Dingshuo Liu ,&nbsp;Daoliang Li ,&nbsp;Qingling Duan","doi":"10.1016/j.compag.2025.110999","DOIUrl":"10.1016/j.compag.2025.110999","url":null,"abstract":"<div><div>Accurate recognition of fish swimming behavior helps in health assessment and disease prevention. However, behavior similarity, drastic pose variations, and class imbalance pose challenges for efficient fish swimming behavior recognition models. Hence, this paper proposes a two-stage framework for recognizing fish swimming behaviors. In the initial stage, the fish skeleton is defined and a novel Skeleton-based Graph Convolutional Network (SGCN) is proposed to extract spatiotemporal features of fish movement. It reliably extracts the positions and spatial relationships of joints, bones, joint motion, and bone motion during fish swimming and reduces data complexity and noise. In the subsequent stage, aiming to improve the performance of the swimming behavior recognition model further, a Hybrid Ensemble (HE) method is designed. This method integrates multiple models’ strengths, reduces individual models’ prediction bias, and improves prediction robustness. To validate the effectiveness of EMGCM, experimental evaluations on a comprehensive PL-behavior dataset achieved an accuracy of 90.31 %, an F1 score of 81.33 %, and a precision of 87.08 %. These results demonstrate that EMGCM outperforms existing methods and supports swimming behavior monitoring in practical aquaculture applications.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110999"},"PeriodicalIF":8.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145105500","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}