Computers and Electronics in Agriculture最新文献

{"title":"Enhancing crop growth forecasting by incorporating estimated uncertainties for time-series hyperspectral data and crop model GECROS simulations into Ensemble Kalman Filter","authors":"Dong Wang ,&nbsp;Paul C. Struik ,&nbsp;Lei Liang ,&nbsp;Xinyou Yin","doi":"10.1016/j.compag.2025.111057","DOIUrl":"10.1016/j.compag.2025.111057","url":null,"abstract":"<div><div>Crop status forecasting by crop model simulations can benefit from assimilating remote sensing observations. When conducting data assimilation (DA) using a common procedure – the Ensemble Kalman Filter (EnKF), arbitrary inflation factors are normally adopted to account for unspecified uncertainties, so as to alleviate filter divergence. Here, we developed a more effective Bayesian methodology, in which the uncertainties were systematically quantified by combining multiple methods in one framework. Its applicability and performance in the EnKF were tested using the crop model GECROS (Genotype-by-Environment interaction on CROp growth Simulator) and the data collected from two years of field experiments for rice. Aboveground biomass (<em>W</em>_above), grain weight (<em>W</em>_grains), aboveground nitrogen (N) content (<em>N</em>_above), grain N content (<em>N</em>_grains) and leaf traits like leaf dry weight, leaf N content and leaf area index were measured in the experiments. Using only the observations from the first year, the uncertain parameters in GECROS were calibrated by a Markov Chain Monte Carlo approach, while the parameters in the uncertainty model that describes the errors of crop model simulations were estimated simultaneously. The calibrated model parameters performed well in the validation year, except for the simulated leaf traits (Normalized Root Mean Squared Error (<em>NRMSE</em>) &gt; 0.38). Remotely sensed leaf traits predicted by a Gaussian Process Regression (GPR) model were more accurate (<em>NRMSE</em> &lt; 0.32), with uncertainties of the remote sensing observations estimated from the GPR model itself. Assimilating simulated and predicted leaf traits with their estimated uncertainties into EnKF prevented filter divergence, and the forecast accuracy of crop model improved in the validation year. Compared with simulation without assimilating in-season remote sensing observations, the assimilation procedure led the <em>NRMSE</em> to decrease from 0.37 to 0.20 for whole-season <em>W</em>_above and <em>N</em>_above and from 0.39 to 0.20 for the end-season <em>W</em>_grains and <em>N</em>_grains. The updated crop traits of our method also agreed better with the measurements than those of common EnKF with arbitrarily assumed uncertainties and with adjusted inflation factors. The developed method contributes to systematic uncertainty analysis in DA and accurate forecasting of crop growth and yield for smart farming.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111057"},"PeriodicalIF":8.9,"publicationDate":"2025-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145227235","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":"Multispectral image reconstruction from RGB image for maize growth status monitoring based on window-adaptive spatial-spectral attention transformer","authors":"Di Song ,&nbsp;Hong Sun ,&nbsp;Esther Ngumbi ,&nbsp;Mohammed Kamruzzaman","doi":"10.1016/j.compag.2025.111062","DOIUrl":"10.1016/j.compag.2025.111062","url":null,"abstract":"<div><div>Multispectral image analysis is an effective way to detect crop growth status. However, the complexity of manufacturing process and technology of multispectral image acquisition equipment make data acquisition expensive. Therefore, a method based on a window-adaptive spatial-spectral attention transformer is proposed to reconstruct multispectral images using RGB images of maize. First, RGB and hyperspectral images of the maize are obtained, and the reflectance data from classic and preferred band combinations are extracted from the hyperspectral image. Then, a transformer model is constructed to evaluate and compare the reconstruction efficacy of the 5-band and 10-band combinations across four attention modes: spatial, spectral, spatial-spectral, and window-adaptive spatial-spectral attention. The best-performing reconstruction results are selected and compared with the original data from three perspectives: image, spectrum, and model effect. The 10-band multispectral image reconstructed by the window-adaptive spatial-spectral attention mechanism is highly similar to the original image, with a reflectance correlation exceeding 0.99. Furthermore, its application in monitoring crop growth status (i.e., maize chlorophyll) yields results closely aligned with actual reflectance data: R_C² is 0.76, R_V² is 0.64, while RMSE_C and RMSE_V are 3.63 mg/L and 2.94 mg/L, respectively. To further explore the model performance, the new sensitive bands are selected to be reconstructed in the maize V7 stage. The results from the chlorophyll content prediction model are as: R_C² is 0.64, R_V² is 0.60, with RMSE_C and RMSE_V are 5.61 mg/L and 5.62 mg/L, respectively. Therefore, the window-adaptive spatial-spectral attention transformer can accurately reconstruct multispectral images and establish precise growth status monitoring models, providing technical support for low-cost field maize growth detection.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111062"},"PeriodicalIF":8.9,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219549","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":"Performance analysis on root-soil separation of Panax notoginseng using DEM-CFD method and air-jet system","authors":"Chenglin Wang,&nbsp;Pengfei Li,&nbsp;Zedong Zhao,&nbsp;Weiyu Pan,&nbsp;Qiyu Han,&nbsp;Zhi He,&nbsp;Maoling Yang,&nbsp;Zhaoguo Zhang","doi":"10.1016/j.compag.2025.111050","DOIUrl":"10.1016/j.compag.2025.111050","url":null,"abstract":"<div><div>PN (Panax notoginseng root) is a kind of Chinese traditional medicinal crop. A large amount of soil is attached to PN after the harvest. Therefore, PN-S (PN and soil mixture) separation operation is a key step in the postharvest processing of PN. However, the root and soil separation operation of PN suffers from two problems: poor root-soil separation effect and unclear separation process. In this study, a method for PN-S separation using jetting is proposed. The three-dimensional scales and geometries of PN and soil particles were measured and calibrated. A coupled fluid–solid separation model of PN-S was constructed using DEM-CFD (discrete element coupling method and computational fluid dynamics). Based on the structure of the air-jet soil separation device and the problems to be solved in the study, a new evaluation standard combining φ (soil separation rate) and η (soil attachment rate) was proposed to quantify the working effect of the air-jet device. This study established experimental platforms for PN and soil air-jet separation devices in both simulated and actual environments. Through the mutual corroboration of physical and simulation experiments, the results prove that the method is feasible. Obtain a clear process of PN and soil separation through the use of a simulation experimental platform. The optimum parameters of the air-jet device were further obtained through the BBD (Box-Behnken Design) method. The optimal parameters are: pressure of 2285 Pa, distance of 10 cm, angle of 90°, and φ of 62.06 %. The fluid–solid coupling model and the new evaluation criteria established in this study provide a theoretical basis for the effective separation of PN-S. The proposed air-jet separation method can effectively separate PN-S. The results of this study provide new ideas and technical support for the optimization and design of PN-S separation equipment.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111050"},"PeriodicalIF":8.9,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219574","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":"Bit-STED: A lightweight transformer for accurate agave counting with UAV imagery","authors":"Diego Villatoro-Geronimo,&nbsp;Gildardo Sanchez-Ante,&nbsp;Luis E. Falcon-Morales","doi":"10.1016/j.compag.2025.111047","DOIUrl":"10.1016/j.compag.2025.111047","url":null,"abstract":"<div><div>This paper presented Bit-STED, a novel and simplified transformer encoder architecture for efficient agave plant detection and accurate counting using unmanned aerial vehicle (UAV) imagery. Addressing the critical need for accessible and cost-efficient solutions in agricultural monitoring, this approach automates a process that is typically time-consuming, labor-intensive, and prone to human error in manual practices. The Bit-STED model features a lightweight transformer design that incorporates innovative techniques for efficient feature extraction, model compression through quantization, and shape-aware object localization using circular bounding boxes for the roughly circular shape of the agave rosettes. To complement the detection model, a novel counting algorithm was developed to manage plants spanning multiple image tiles accurately. The experimental results demonstrated that the Bit-STED model outperformed the baseline models in terms of detection and agave plant count performance. Specifically, the Bit-STED nano model achieved F1 scores of 96.66% on a map with younger plants and 96.43% on a map with larger, highly overlapping plants. These scores surpassed state-of-the-art baselines, such as YOLOv8 Nano (F1 scores of 96.42% and 96.38%, respectively) and DETR (F1 scores of 93.03% and 85.61%, respectively). Furthermore, the Bit-STED nano model was significantly smaller, being less than one-eighth the size of the YOLOv8 nano model (1.4 MB compared to 12.0 MB), had fewer trainable parameters (0.35M compared to 3.01M), and was faster in average inference times (14.62 ms compared to 18.28 ms).</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111047"},"PeriodicalIF":8.9,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219576","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":"Design and ground testing of a Zero-Discharge plant growth system for microgravity Applications","authors":"Claudia Garrido-Ruiz ,&nbsp;Juan D. González-Teruel ,&nbsp;Chihiro Dixon ,&nbsp;Sarah Schreck ,&nbsp;Curtis Bingham ,&nbsp;Thane Winward ,&nbsp;Riley Hutchings ,&nbsp;Gail E. Bingham ,&nbsp;Bruce Bugbee ,&nbsp;Scott B. Jones","doi":"10.1016/j.compag.2025.111044","DOIUrl":"10.1016/j.compag.2025.111044","url":null,"abstract":"<div><div>Plant-based bioregenerative life-support systems play an essential role for long-duration space missions, offering a renewable source of fresh, nutritious food and psychological benefits for crew members. As human space missions extend further and last longer, developing efficient technologies for space agriculture becomes increasingly critical. In microgravity, altered fluid dynamics change water, nutrient, and gas distribution within the root-zone, potentially limiting plant growth. The Utah Reusable Root Module (URRM) system housed within NASA’s Ohalo III Crop Production System addresses these challenges through five root modules equipped with automated fertigation, redundant moisture sensors, and media containment materials. Designed to support repetitive harvests of pick-and-eat vegetables with minimal crew intervention, the URRM advances water and nutrient management strategies for space-based agriculture. Preliminary ground tests with Mizuna (Brassica rapa var. nipposinica) demonstrated that the URRM maintained optimal root-zone conditions and uniform resource distribution, yielding more than 1 kg of fresh biomass over 17 days. The use of different top cover designs and materials across root modules affected plant establishment and yield, as well as evapotranspiration, whereas water use efficiency (WUE) exceeded 2 g L^-1 across the system. These findings highlight the URRM’s ability to support crop production using automated state-of-the-art technologies, strengthening the feasibility of long-term human space exploration.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111044"},"PeriodicalIF":8.9,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219552","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":"CATR-Net: Cattle–Attentive transformer with adaptive and enhanced segmentation and recognition","authors":"Xiaopu Feng ,&nbsp;Jiaying Zhang ,&nbsp;Yongsheng Qi ,&nbsp;Liqiang Liu ,&nbsp;Yongting Li","doi":"10.1016/j.compag.2025.111038","DOIUrl":"10.1016/j.compag.2025.111038","url":null,"abstract":"<div><div>In open–range cattle–face analysis, conventional segmentation networks struggle to preserve fine–scale edge cues, and recognition networks weakly model salient regions and local–global context, yielding brittle performance under changing poses and illumination. We propose CATR–Net, an end–to–end framework that unifies segmentation and identification. In its segmentation branch, a Multiscale Edge–enhanced Upsampling Module (MEUM) is grafted onto the DA–TransUNet decoder to restore high–frequency boundaries and suppress up–sampling blur; in the recognition branch, a Dynamic Contextual Attention Module (DCAM) is inserted between the Stem and MaxViT blocks, and Dynamic Adaptive Interaction Normalization (DAIN) replaces the static Layer Normalization in LSRA (Local Region Self-Attention) with DyT (Dynamic tanh), together enabling pose– and scale–aware fusion of local priors with global dependencies. The recognition loss is further equipped with a confidence–gap regularizer that dynamically tunes the Dynamic–Tanh parameters, amplifying ambiguous features while stabilizing high–confidence activations. On a 57645–image multi–pose dataset, the segmentation branch achieves 93.35 % mIoU and 96.45 % mDSC with a 417  MB model, whereas the recognition branch attains 97.03 % accuracy and 95.19 % F1-score with a 457  MB footprint—both surpassing state–of–the–art baselines at comparable complexity.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111038"},"PeriodicalIF":8.9,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219577","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 derivative approach for efficient hydroponic vertical farm monitoring using hyperspectral vision","authors":"Maria Merin Antony ,&nbsp;M.M. Bijeesh ,&nbsp;C.S. Suchand Sandeep ,&nbsp;Murukeshan Vadakke Matham","doi":"10.1016/j.compag.2025.111029","DOIUrl":"10.1016/j.compag.2025.111029","url":null,"abstract":"<div><div>Vertical indoor hydroponic farms offer sustainable solutions in land scarce countries to foster agriculture productivity for addressing growing demand. Such farms require extensive controllability of the growing conditions to ensure year round-cultivation of diverse crops within the space available. Continuous monitoring of the crops and early remedial measures are essential to ensure non-compromised, high-quality yield from these farms. Currently, most farms rely on human vision based monitoring, which is quite subjective and time-consuming and could be ineffective in identifying crop stresses at early stages. Hence, efficient management of these farms requires advanced automated systems to monitor crop health, including possible stress factors such as nutrient, water, and light deficiencies at early stages to enable timely intervention. This research, in this context, explores innovative strategies using assessment parameters such as spectral ratios and derivative reflectance derived from hyperspectral images for crop monitoring. Customized spectral index for nutrient deficiency detection and approaches for quantification of derivative spectra for stress detection are developed. These strategies can be used to rapidly detect the stresses at the early stages non-destructively (within hours in case of light and water deficiencies) and could promptly guide in timely remedial actions. The proposed method offers automation possibilities for non-invasive monitoring systems utilizing hyperspectral vision. This non-invasive imaging system integrated on a robotic platform is envisaged to revolutionize the development of unmanned indoor hydroponic farms for a sustainable future.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111029"},"PeriodicalIF":8.9,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219550","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":"Feeding behavior recognition of group-housed pigs based on pose estimation and keypoint features discrimination","authors":"Yuhang Hu ,&nbsp;Xin Dai ,&nbsp;Baisheng Dai ,&nbsp;Ran Li ,&nbsp;Junlong Fang ,&nbsp;Yanling Yin ,&nbsp;Honggui Liu ,&nbsp;Weizheng Shen","doi":"10.1016/j.compag.2025.111039","DOIUrl":"10.1016/j.compag.2025.111039","url":null,"abstract":"<div><div>In the field of intelligent sensing for smart animal husbandry, accurate recognition of feeding behavior in group-housed pigs is crucial for achieving precision farming and improving pig welfare. Currently, pig feeding behavior recognition relies on detection boxes-based methods, which are difficult to exclude Non-Nutritive Visiting Behavior within the feeding zone. To precisely recognize the feeding behavior of group-housed pigs, this study proposes a feeding behavior recognition method based on pose estimation and keypoint features discrimination. Firstly, Pig-HRNet is designed to estimate the pose of group-housed pigs, in which a Context Transformer (COT) attention module is specially introduced to detect the keypoints of pigs more accurately under crowded conditions. Secondly, by analyzing the correlation between keypoints and feeding zone, group-housed pigs are divided into visiting the feeding zone and Non-Feeding Behavior (NFB). For visiting the feeding zone, the behaviors are further categorized into Feeding Behavior (FB) and Non-Nutritive Visiting Behavior (NNVB). The experimental data of group-housed pigs were collected in commercial pig farms, including a total of 1400 video frames. Experimental results show that the Pig-HRNet model achieves an average precision (AP) of 97.1% in estimating pig poses. Compared to other pose estimation network models such as KAPAO, HigherHRNet, DeepLabCut, and HRNet, the detection AP improved by 69.0%, 16.3%, 12.3%, and 0.5%, respectively. The feeding behavior recognition method proposed in this paper achieved precision and recall rates of 98.8% and 99.9%, respectively. The relevant results indicate that the proposed feeding behavior recognition method performs well, while also meeting the requirement for accurately estimating pig poses under crowded conditions. The feeding behavior dataset established in this paper has been shared on <span><span><u>https://github.com/IPCLab-NEAU/Group-housed-pigs-Feeding-Behavior-Recognition</u></span><svg><path></path></svg></span> for use by the precision animal husbandry research community.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111039"},"PeriodicalIF":8.9,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219551","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":"DEM simulation of straw-soil-cutting system interactions for performance assessment of a spin-descent straw cutter","authors":"Rongrong Li ,&nbsp;Hongwen Li ,&nbsp;Shuofei Yang ,&nbsp;Caiyun Lu ,&nbsp;Zhengyang Wu ,&nbsp;Zhinan Wang","doi":"10.1016/j.compag.2025.111022","DOIUrl":"10.1016/j.compag.2025.111022","url":null,"abstract":"<div><div>Accurate measurement of straw mulch mass requires effective separation of residues inside and outside the sampling frame. To achieve this, a novel spin-descent straw cutter was designed, integrating a stabilizing ring and a rotating and descending separation column. The stabilizing ring pressed the straw to reduce movement, while the separation column simultaneously rotated and descended to cut and separate the straw at the frame boundary, representing an integrated structure not previously reported. A discrete element method (DEM) model was established to simulate straw-soil-cutting system interactions, and three blade types (sawtooth, corrugated, and notch-shaped) were compared. Response surface methodology and multi-objective optimization were employed to balance cutting rate, power consumption, and soil disturbance area. Bench tests confirmed the DEM predictions with errors within 12 %. Results demonstrated that the sawtooth blade achieved the most favorable trade-off, with optimal parameters of 287 N downward force, 230 rpm rotation speed, and 10 mm insertion depth. The study highlighted the unique suitability of the spin-descent cutter for automated straw mulching detection and provided a validated simulation–optimization framework to inform future cutter designs and broader applications in conservation tillage.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111022"},"PeriodicalIF":8.9,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219631","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":"Intelligent dual-modal hyperspectral imaging system with deep learning for in-field soil fauna identification","authors":"Jing Luo ,&nbsp;He Zhu ,&nbsp;Ronggui Tang ,&nbsp;Sailing He","doi":"10.1016/j.compag.2025.111040","DOIUrl":"10.1016/j.compag.2025.111040","url":null,"abstract":"<div><div>Accurate and efficient identification of living soil fauna is crucial for ecological biodiversity assessment and sustainable agriculture practices, yet traditional methods are labor-intensive, time-consuming, and often ineffective in complex field environments. This study pioneeres a novel dual-modal hyperspectral soil fauna identification (HSFI) system, which ingeniously Integrates both reflectance and fluorescence hyperspectral imaging with a custom-designed deep learning model, HSFI-Net, for robust semantic segmentation. This synergistic approach effectively overcomes the challenges of rapidly and accurately identifying soil fauna, even those partially concealed with heterogeneous soil backgrounds. Using the HSFI system, we established a comprehensive dual-modal spectral database for diverse soil macrofauna, including earthworms, centipedes, scorpions, pillworms, millipedes, crickets, beetles, ants and so on. Extensive experimental evaluations demonstrated the system’s exceptional robustness and high precision, achieving average IoU of 0.734 across various exposure levels and various species under challenging soil conditions. Furthermore, in-field experiments successfully validated the system’s capability for in-situ identification and analysis of soil fauna’s horizontal and vertical distribution. This innovative HSFI system offers an automated, intelligent tool for monitoring soil biodiversity, which is vital for precision agricultural management, environmental conservation, and understanding the intricate dynamics of soil ecosystems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111040"},"PeriodicalIF":8.9,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219633","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}