Computers and Electronics in Agriculture最新文献

{"title":"Cross-domain adversarial learning for forage mapping and alpha-diversity assessment from UAV hyperspectral imagery in desert rangelands","authors":"Tao Zhang ,&nbsp;Chuanzhong Xuan ,&nbsp;Zhaohui Tang ,&nbsp;Xinyu Gao ,&nbsp;Fei Cheng ,&nbsp;Suhui Liu","doi":"10.1016/j.compag.2025.111001","DOIUrl":"10.1016/j.compag.2025.111001","url":null,"abstract":"<div><div>Accurate assessment of forage distribution and diversity is crucial for sustainable grazing management in desert rangelands. Unmanned aerial vehicle (UAV)-based hyperspectral imagery provides a robust data source for the fine-scale monitoring of forage resources. However, the complex nature of grassland environments, combined with the high cost and time-consuming process of sampling, results in a scarcity of labeled samples, which limits the representational ability of models for downstream tasks. To address these challenges, this study proposes the domain-adversarial multi-view contrastive network (DA-MVCNet) for forage mapping and alpha (<span><math><mi>α</mi></math></span>)-diversity assessment in desert rangelands. The network incorporates a domain adversarial strategy to mitigate discrepancies in feature distribution between the source and target domains, and employs a multi-view supervised contrastive learning module to promote feature alignment, thereby enhancing feature discrimination in the target domain. Experimental results demonstrate that DA-MVCNet achieves an overall accuracy of 92.52% in forage classification, outperforming other state-of-the-art methods, while maintaining low computational complexity with only 0.11 G FLOPs and 0.30 M parameters. Furthermore, we mapped the spatial distributions of <span><math><mi>α</mi></math></span>-diversity indices – including species richness, the Shannon–Wiener index, the Simpson index, and Pielou’s evenness index – using a grid-based approach. Results indicate that with increasing grazing pressure, all four diversity indices significantly decreased (<span><math><mrow><mi>P</mi><mo>&lt;</mo><mn>0</mn><mo>.</mo><mn>05</mn></mrow></math></span>), effectively revealing the impact of grazing disturbance on community structure. This study provides a new pathway for forage mapping and diversity assessment via UAV hyperspectral remote sensing, offering technical support for the intelligent monitoring and management of rangeland ecosystems. The code is available at <span><span>https://github.com/zhang2508/DA-MVCNet</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111001"},"PeriodicalIF":8.9,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219689","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 super-hydrophobic tactile sensor for damage-free fruit grasping","authors":"Zhen Wei ,&nbsp;Chunjiang Zhao ,&nbsp;Yuehua Huang ,&nbsp;Xinglan Fu ,&nbsp;Jing Li ,&nbsp;Guanglin Li","doi":"10.1016/j.compag.2025.111043","DOIUrl":"10.1016/j.compag.2025.111043","url":null,"abstract":"<div><div>As a crucial component of smart agriculture, fruit-picking robots are widely applied in agricultural production. However, the rigid materials and imprecise force control of conventional robotic end-effectors often result in significant fruit damage during harvesting operations. To address this limitation, we developed a novel super- hydrophobic tactile sensor integrated with a real-time signal acquisition and control system based on Arduino microcontroller and Python programming. This integrated sensor-end effector system achieves damage-free fruit grasping capabilities through real-time monitoring and precise control of grasping forces during robotic operations. The developed sensor exhibits exceptional performance metrics, including a high sensitivity of 5.57 kPa⁻¹, a broad detection range (0.01–450 kPa), and remarkable durability exceeding 8000 operational cycles, with rapid response characteristics (103 ms response time and 73 ms recovery time). Furthermore, the sensor’s super-hydrophobic properties enable reliable operation across a wide humidity range (20–80 % RH) and even in underwater environments. This research provides valuable insights and technical guidance for developing damage-free agricultural robotic systems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111043"},"PeriodicalIF":8.9,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219691","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":"Time-optimal and path-optimal motion planning for selective broccoli harvesting manipulator","authors":"Shuo Kang ,&nbsp;Dongfang Li ,&nbsp;Sifang Long ,&nbsp;Yongkai Ye ,&nbsp;Fuming Kuang ,&nbsp;Dongdong Du ,&nbsp;Jun Wang","doi":"10.1016/j.compag.2025.111007","DOIUrl":"10.1016/j.compag.2025.111007","url":null,"abstract":"<div><div>Due to the asynchronous maturation characteristics of broccoli, manual harvesting in multiple batches encounters labor shortages. Consequently, there is an urgent requirement for the development of a selective broccoli harvesting robot. One of the crucial technologies in the broccoli selective harvesting robot is the motion planning algorithm, which quickly generates safe and feasible paths to avoid collisions, ensuring efficient completion of the harvesting task. However, existing motion planning algorithms struggle to balance planning speed, path quality, and success rate simultaneously. The desired motion planning algorithm for the robotic arm should enable continuous, smooth, safe, and fast movements, resembling the motion of a human arm, as it travels between the cutting and collection points, thereby enhancing the efficiency of broccoli harvesting. To address this issue, we proposed flexible edge checking and uniform sampling strategies to enhance the sampling and search processes, thereby expediting planning speed. Additionally, a Markov optimiser was incorporated to optimise path length and smoothness. Our proposed algorithm is named Time-Optimal and Path-Optimal (TOPO), which maintains and enhances the convergence properties and asymptotic optimality of Batch Informed Trees (BIT*), enabling the manipulator to exhibit smooth, efficient, and adaptive trajectories. To verify the effectiveness of our proposed algorithm, we conducted multi-dimensional collaborative simulation experiments and prototype verification experiments. The results demonstrate that compared with Rapidly-exploring Random Tree Star (RRT*), Rapidly-exploring Random Tree Connect (RRT-Connect), Fast Marching Tree Star (FMT*), Adaptively Informed Trees (AIT*), Advanced BIT*(ABIT*) and BIT* algorithms, the TOPO algorithm can reduce path planning time by 50%-73%, shorten path length by 4%-11%, minimise operating time by 9%–32%, and improve planning success rate to 97%. It can reduce the operating time of harvesting a single broccoli to within 11 s, thus enabling the selective harvesting robot to perform tasks more efficiently.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111007"},"PeriodicalIF":8.9,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145219688","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":"Dynamic weed control using selective laser application with object tracking and target scheduling","authors":"Juan Ignacio Vargas Fernández ,&nbsp;Sam Wane ,&nbsp;Tito Arevalo-Ramirez ,&nbsp;Fernando Auat Cheein","doi":"10.1016/j.compag.2025.111004","DOIUrl":"10.1016/j.compag.2025.111004","url":null,"abstract":"<div><div>Selective laser application for weed control is emerging as one of the most sustainable practices for various crops. The system targets weeds using a laser beam with specific time and intensity settings to eliminate undesired plants through thermal damage. However, this process — commonly known as static weed laser treatment — reduces machinery efficiency, as the platform must remain stationary until all visible weeds are treated. To address this limitation, the current work proposes a dynamic laser weeding approach that predicts weed positions while the platform is in motion, thereby improving operational efficiency. Several deep learning architectures (e.g., YOLO series for weed detection and DeepSORT for weed tracking) are evaluated to identify the most effective models for detecting and tracking multiple weeds in RGB images. In addition, a time-constrained scheduling strategy is implemented to determine the order in which weeds are treated, minimizing the number of missed targets. We find that receding horizon control offers the best performance, particularly under strict time and energy constraints. Field deployment results show that YOLOv7 achieves the highest precision, recall, and mean Average Precision (mAP) for weed detection. The dynamic laser weeding system significantly outperforms the static counterpart, enabling up to 2.8 times faster movement while successfully treating 90% of detected weeds. This work presents a proof of concept for dynamic weeding, laying the foundation for future developments in intelligent, autonomous crop protection systems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111004"},"PeriodicalIF":8.9,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158231","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":"Evaluating the potential of very high-resolution satellite data for the enhanced estimation of rice aboveground biomass by combining spectral and spatial information","authors":"Tianyue Xu ,&nbsp;Fumin Wang ,&nbsp;Zhou Shi ,&nbsp;Marc Peaucelle ,&nbsp;Jean-Pierre Wigneron","doi":"10.1016/j.compag.2025.110997","DOIUrl":"10.1016/j.compag.2025.110997","url":null,"abstract":"<div><div>Monitoring aboveground biomass (AGB) using high spatial and temporal resolution remote sensing data is important for smart agriculture. Significant technological advances have been made in developing satellites with very high spatial resolution, delivering a promising avenue for vegetation observations. However, the high costs and limited revisit periods of high-resolution satellites hinder their widespread use, leaving the feasibility of combining vegetation indices (VIs) and textures derived from satellite images for AGB estimation uncertain and the quantitative improvements achieved by incorporating textures into estimation unclear. Airborne hyperspectral imaging with high spectral and spatial resolution offers a fresh opportunity to simulate the satellite imaging process objectively and realistically across both spectral and spatial dimensions. The study first evaluated the potential benefits of combining textures and VIs derived from different high-resolution satellites to enhance AGB retrieval. Rice samples and UAV hyperspectral data were collected throughout the rice growth cycle over three consecutive years. Each hyperspectral image was resampled in spectral and spatial dimensions to simulate nine multispectral satellites with sub-meter spatial resolution (WorldView-3, WorldView-2, GeoEye-1, SuperView-1C, GaoFen-2, Beijing-2, Jilin-1, GeoSat-2, KomPast-2). VIs, textures, and their combinations were employed to establish AGB models for the pre-heading, post-heading, and the entire growth stage, respectively. The results showed that combining VIs and textures always achieved the greatest rice AGB estimations, with the integration of multiple satellite data always yielding the best outcomes (overall validation rRMSE ≤ 0.35). For the texture-based monitoring, the impact of satellite spatial resolution was more pronounced on influencing the estimation effectiveness than spectral bands. The monitoring accuracy of rice AGB demonstrated a nonlinear decreasing trend as the spatial resolution dropped, and combining VIs and textures mitigated the negative impact of reduced spatial resolution on the monitoring accuracy of rice AGB. The combination of VIs and textures showed a compensatory effect and combining VIs and textures derived from red-edge band could offset the impact of the reduced spatial resolution on AGB estimation. The involvement of textures in modelling exerted an overall bigger impact on rice AGB estimation than the inclusion of red-edge variables. Satellites with higher spatial resolution and a red-edge band always performed the best in AGB estimation. This study facilitates the optimization of sensor design and farmland management.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110997"},"PeriodicalIF":8.9,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158071","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 UNet-GAN two-stage network for rapid and accurate prediction of apple optical properties from single multi-frequency images","authors":"Dong Hu ,&nbsp;Yumeng Peng ,&nbsp;Tianze Jia ,&nbsp;Zhizhong Sun ,&nbsp;Chang Zhang ,&nbsp;Guoquan Zhou","doi":"10.1016/j.compag.2025.111028","DOIUrl":"10.1016/j.compag.2025.111028","url":null,"abstract":"<div><div>Spatial frequency domain imaging (SFDI) is a non-invasive optical imaging technique widely used for the quantitative determination of fruit tissue optical properties, specifically absorption coefficient (<em>μ_a</em>) and reduced scattering coefficient (<em>μ_s^’</em>). However, traditional SFDI methods rely on multiple frequency and phase images, limiting real-time imaging capabilities. To address this issue, we present a novel rapid prediction method based on a two-stage deep neural network architecture, termed <strong>FSGOP</strong> (<strong>F</strong>requency-<strong>S</strong>patial Attention UNet and <strong>G</strong>AN-based two-stage network for <strong>o</strong>ptical <strong>p</strong>roperties prediction). Compared with conventional three-phase demodulation SFDI, this method reduces the acquisition time by approximately 5/6 and requires only 0.21 s for inference. In the first stage, a UNet network enhanced by Frequency-Spatial Attention (FSA) is employed to effectively decouple the multi-frequency components. In the second stage, a Generative Adversarial Network (GAN) is utilized to predict the optical properties, thereby enabling the simultaneous extraction of <em>μ_a</em> and <em>μ_s^’</em> maps under different frequency conditions from a single multi-frequency mixed fringe image. In experiments on apples, pears, and peaches, the method yielded normalized mean absolute errors of 0.10 (<em>f₁</em>) and 0.09 (<em>f₂</em>) for <em>μ_s^’</em>, and 0.07 and 0.06 for <em>μ_a</em>, respectively. The results revealed significant complementary information in the optical property maps at different frequencies, with lower frequencies being more sensitive to subsurface damage and higher frequencies revealing surface texture features more effectively. This method enhances information utilization and real-time performance in multi-frequency imaging, offering a rapid, accurate, and low-cost solution for optical property extraction and quality inspection of agricultural products.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111028"},"PeriodicalIF":8.9,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158072","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":"Salt stress estimation in pumpkin germplasm based on maximum likelihood statistical modeling of the leaf color space distribution","authors":"Zhang Xinyan ,&nbsp;Liu Quan ,&nbsp;Li Shuai ,&nbsp;Feng Huaxiong ,&nbsp;Luo Liang ,&nbsp;Tan Zuojun ,&nbsp;Shen Huan ,&nbsp;Bie Zhilong ,&nbsp;Xie Jing","doi":"10.1016/j.compag.2025.110982","DOIUrl":"10.1016/j.compag.2025.110982","url":null,"abstract":"<div><div>Breeding salt-tolerant pumpkin cultivars is crucial for improving crop quality and yield. In this study, a high-resolution imaging-based phenotyping platform was developed to capture true leaf images of pumpkin seedlings subjected to salt stress, and plant experts conducted field assessments to determine the severity of salt damage. After image preprocessing, binary mask images were generated, and the maximum likelihood values of normalized intensity were extracted in the red, green, and blue channels to establish a salt stress status index (β) for characterizing stress levels. The β value shows a strong correlation with SPAD value,<!--> <!-->which indicates that it can effectively reflect the chlorophyll content in leaves, thereby reflecting the physiological changes in leaves affected by salt stress. A leaf texture factor (α) was employed to investigate the directional characteristics of the leaf texture, it can facilitate the effective differentiation of the clusters identified in the clustering analysis and enhance model precision by incorporating detailed leaf structural features. The performance of machine learning, deep learning, and statistical modeling approaches was compared. Statistical model integrating β and α exhibited superior predictive accuracy, with a coefficient of determination, root mean square error, and mean absolute error of 0.901, 0.057, and 0.046, respectively, in the validation dataset. Accuracy assessment among 49 germplasm accessions achieved 95.65 %, demonstrating the model’s reliability. Compared to conventional salt injury assessment, this approach offers higher efficiency and greater objectivity, enabling rapid and accurate identification of salt stress levels in pumpkin seedlings. This study provides a rapid and efficient method for assessing salt stress in pumpkin seedlings, contributing to a deeper understanding of stress response mechanisms and facilitating the selection of salt-tolerant cultivars. Moreover, these findings offer a valuable reference for salt stress identification in other plant species.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 110982"},"PeriodicalIF":8.9,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158228","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":"Data-driven localization of the TOMGRO model: Cultivar-specific parameter optimization for Shanghai greenhouse tomato production","authors":"Yue Sun ,&nbsp;Qingqing Ju ,&nbsp;Yiyang Li ,&nbsp;Linyi Li ,&nbsp;Yuhang Wang ,&nbsp;Juan Yang ,&nbsp;Tingting Qian","doi":"10.1016/j.compag.2025.111025","DOIUrl":"10.1016/j.compag.2025.111025","url":null,"abstract":"<div><div>Crop models are an integral component in greenhouse control systems, enabling the simulation of plant responses to environmental conditions and facilitating optimal operational decisions for high productivity with low energy use. However, existing crop models often lack transferability beyond their original development conditions. Additionally, cultivar-specific parameterization remains challenging, as some parameters can be empirically determined while others require complex calibration. This study adapted the reduced TOMGRO model to simulate growth and yield for four local tomato cultivars under Shanghai greenhouse conditions. Through Sobol’s global sensitivity analysis and Bayesian optimization, four highly influential parameters were identified and optimized, including growth efficiency (E), maintenance respiration coefficient (r_m), extinction light coefficient (K), and leaf quantum efficiency (Q_e). This combined approach provides an effective framework for model calibration, with the calibrated model achieving an average R² &gt; 0.94 for node number, plant dry weight, fruit dry weight, and leaf area index predictions in all cultivars. Model validation using 2023–2024 greenhouse data confirmed model effectiveness for the target variables (average R² &gt; 0.92 for cultivar QX and &gt; 0.88 for LZ), whereas the model showed limitations in simulating mature fruit growth. This calibrated model offers reliable predictions of key growth variables, informing both plant breeding and greenhouse management.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111025"},"PeriodicalIF":8.9,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158230","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":"PWD-lightweight and feature fusion network for multi-stage joint detection of pine wilt disease","authors":"Jie Zhou ,&nbsp;Fang Wang ,&nbsp;Hongping Zhou ,&nbsp;Haifeng Lin","doi":"10.1016/j.compag.2025.111015","DOIUrl":"10.1016/j.compag.2025.111015","url":null,"abstract":"<div><div>Pine Wilt Disease (PWD) has caused irreversible damage to the health of pine forests around the world. Accurate detection and identification is the prerequisite for taking measures to prevent the spread of PWD. To tackle this challenge, this study presents PWD-Lightweight and Feature Fusion Network (PWD-LWFFNet), a specially designed object detection model for detecting PWD in pine trees. PWD-LWFFNet uses lightweight EIBNet as its backbone network, which is mainly implemented by the EIB module for lightweighting. The EIB module utilizes the inverted bottleneck module and incorporates Extra Depthwise convolution (DW) to minimize the number of parameters while ensuring computational efficiency. In the neck network, PAFPN-4Net was designed as a multi-scale feature fusion network and an FMN module was introduced. The FMN module merges the input features with the global information through its ‘aggregation’ and ‘modulation’ components. This setup allows the network to dynamically focus its attention on the minute disease details in the early stages of pine tree infection. Four detection heads are designed and integrated with the Enhanced Multi-scale Attention (EMA) mechanism to capture fine-grained features. Finally, PIoUv2 is selected as the loss function to guide anchor boxes along the optimal path for regression. Comprehensive experiments demonstrate that PWD-LWFFNet exhibits excellent performance in detecting PWD, which the mean Average Precision (mAP) is 94.1%. It particularly excels in detecting small, early-stage targets compared to other mainstream models, with an Average Precision (AP) of 83.4%. The detection accuracies for the middle, late, and tree mortality stages reach 97.3%, 98.4%, and 97.2% respectively. When compared with existing mainstream models, PWD-LWFFNet demonstrates state-of-the-art performance. Experiments conducted on the PWD dataset established in this paper show that PWD-LWFFNet maintains good performance even in environments with background noise, validating the effectiveness of the model in early disease detection and management in pine forests. Its lightweight design provides a guarantee for practical application deployment.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111015"},"PeriodicalIF":8.9,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158070","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":"Cover damage detection in round cotton modules using convolutional neural networks (CNNs)","authors":"Md Zafar Iqbal ,&nbsp;Robert G. Hardin ,&nbsp;Joshua Peeples ,&nbsp;Edward M. Barnes","doi":"10.1016/j.compag.2025.111023","DOIUrl":"10.1016/j.compag.2025.111023","url":null,"abstract":"<div><div>Plastic contamination in cotton threatens the economic viability and global reputation of US cotton. In the US, most contaminants likely originate from damaged plastic covers on round cotton modules, as loose pieces of cover can be torn and entangled in cotton by handling equipment. This study aimed to develop a robust convolutional neural network (CNN)-based detection model to identify cover damage on modules during handling, enabling necessary interventions to mitigate contamination. To achieve this objective, several models, including two-stage, one-stage, and detection transformers, were trained using images of modules with damaged covers. Following evaluation, the most effective model (YOLOv8l) was further optimized through pruning and fine-tuning, resulting in the proposed YOLOv8-wd model. This model achieved a mean average precision (mAP) of 92 % for detecting module cover damages, with an inference speed of 6.20 ms per image using sparse-aware engine. The proposed model demonstrated comparable accuracy to YOLOv8l while being 62.71 % lighter and 50.40 % faster. Model testing was conducted using images collected by a system installed on a module truck and a loader used for module handling at a fully operational gin and field. The loader handled 1,801 modules, capturing 6,935 images, while the truck handled 2,094 modules, yielding 32,584 images. From these images, YOLOv8-wd identified cover damage in 4.72 % of loader-handled and 3.92 % of truck-handled modules, though actual rates may be higher. Furthermore, using the model, the system provided clear status indicators (cover-damaged or undamaged) and unique ID’s for each module. The findings of this study could be used to reduce economic losses resulting from damaged covers of round cotton modules.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":"239 ","pages":"Article 111023"},"PeriodicalIF":8.9,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158073","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}