[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-875":3,"related-875":64},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":8,"published_at":11,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":23,"tags":25,"primary_topic":32,"secondary_topics":38,"classification":39,"search_phrases":59,"slug":62,"view_count":21,"doi":8,"paper":8,"created_at":63},875,"新疆农业大学低成本快速构建田间棉花3D点云评估冠层性状:UAV航拍141张影像约20分钟重建,效率优于SfM-MVS和Instant-NGP","https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619","新疆农业大学邱浩等学者在《MDPI Agronomy》发表论文,提出低成本快速构建田间棉花3D点云并评估冠层性状的技术堆栈。针对传统3D数据采集成本高、处理时间长、运营效率低等瓶颈,集成快速重建算法与基于地面控制点(GCP)的尺度恢复机制,能够以低成本和短时间从UAV航拍图像生成真实尺度3D点云。仅使用141张UAV影像和约20分钟重建时间即可高效重建两个5.5m×5.5m棉花小区的高质量尺度精确点云,在重建效率和点云完整性方面均显著优于SfM-MVS和Instant-NGP;研究进一步量化分析收获期冠层高度变异系数、孔隙度分布和冠层体积分数,为精准栽培管理与机械化兼容种植提供数据支持。",null,"## 1. Introduction\n\nCotton, as one of the most economically important crops worldwide [[1](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B1-agronomy-16-01619)], not only provides the core natural fiber raw material for the textile industry, but also serves as a significant oilseed and cash crop. Its industrial chain plays an irreplaceable supporting role in global rural employment and economic stability [[2](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B2-agronomy-16-01619),[3](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B3-agronomy-16-01619),[4](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B4-agronomy-16-01619)]. Xinjiang, as the major cotton-producing region in China, holds a strategically irreplaceable role in ensuring the supply security of the national cotton industry [[5](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B5-agronomy-16-01619)]. With the rapid advancement of agricultural mechanization and smart farming technologies, Xinjiang cotton has achieved a high degree of mechanization in key production processes such as sowing and harvesting, greatly improving production management efficiency and reducing labor dependence. However, the current analysis of canopy characteristics for field-grown cotton still relies primarily on traditional manual field survey methods, which suffer from high labor intensity, strong subjective bias in observation, limited sampling coverage, and insufficient spatial representativeness [[6](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B6-agronomy-16-01619),[7](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B7-agronomy-16-01619)], making it difficult to achieve large-scale, refined, and time-series dynamic monitoring and quantitative evaluation of canopy characteristics at the field scale [[8](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B8-agronomy-16-01619),[9](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B9-agronomy-16-01619)]. Therefore, it is necessary to develop efficient and scalable technical approaches for quantitative analysis and comprehensive evaluation of canopy structure and growth status in field-grown cotton populations, thereby providing scientific data support and technical foundations for growth diagnosis, cultivation regulation, and population optimization [[10](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B10-agronomy-16-01619),[11](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B11-agronomy-16-01619)].\n\nTo overcome the limitations of the aforementioned manual surveys, the rapid development of UAV remote sensing technology and three-dimensional (3D) reconstruction methods in recent years has brought new technical opportunities for crop canopy analysis. By equipping UAVs with RGB cameras, multispectral sensors, or even LiDAR, researchers can acquire both two-dimensional spectral information and three-dimensional spatial structural information of the canopy, and subsequently estimate key phenotypic parameters such as plant height, leaf area index, biomass, and canopy porosity [[12](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B12-agronomy-16-01619),[13](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B13-agronomy-16-01619),[14](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B14-agronomy-16-01619),[15](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B15-agronomy-16-01619)]. Among these, 3D point cloud data, owing to their ability to faithfully reconstruct the stereoscopic spatial configuration of the canopy, exhibit unique advantages in canopy vertical stratification analysis, light radiation interception assessment, population ventilation and light transmission diagnosis, and planting density optimisation [[16](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B16-agronomy-16-01619)]. In recent years, deep learning-based 3D reconstruction methods have made significant progress, among which Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have become research hotspots in this field [[17](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B17-agronomy-16-01619),[18](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B18-agronomy-16-01619)]. NeRF can generate high-quality, photorealistic 3D reconstructions from sparse-view 2D images, demonstrating unique advantages in fine-scale plant structure modelling. 3DGS, on the other hand, explicitly represents scene geometry via Gaussian primitives, showing remarkable potential in reconstruction efficiency and scalability. These two types of methods have been applied in phenotypic studies of various crops, including wheat [[19](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B19-agronomy-16-01619)] and pepper [[20](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B20-agronomy-16-01619)], and have demonstrated promising application prospects in tasks such as plant organ segmentation, yield-related trait prediction, and large-scale field reconstruction [[21](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B21-agronomy-16-01619)]. However, current mainstream methods for field 3D data acquisition still face significant technical bottlenecks and engineering obstacles. On one hand, although LiDAR systems based on ground or UAV platforms can directly acquire high-precision 3D point clouds, they suffer from high equipment costs, large data volumes, and complex pre-processing workflows [[22](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B22-agronomy-16-01619),[23](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B23-agronomy-16-01619),[24](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B24-agronomy-16-01619)]. Moreover, under field conditions, the multiple-return signals of LiDAR are susceptible to canopy occlusion and wind-induced disturbances, leading to data redundancy and local information loss [[25](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B25-agronomy-16-01619)]. On the other hand, traditional point cloud generation methods based on Structure from Motion (SfM) require a large number of UAV aerial images and high overlap rates—typically hundreds or even thousands of images per experimental plot—resulting in lengthy reconstruction times and heavy computational resource consumption. Furthermore, under complex field illumination, mutual plant occlusion, and repetitive leaf texture conditions, these methods are prone to point cloud voids, excessive noise, and reduced geometric accuracy. These technical bottlenecks make it difficult for existing 3D canopy monitoring methods to achieve a balance among cost, efficiency, and accuracy, and thus they cannot truly meet the practical demands of rapid field monitoring, real-time diagnosis, and large-scale application [[20](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B20-agronomy-16-01619),[26](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B26-agronomy-16-01619),[27](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B27-agronomy-16-01619)].\n\nTo address the above issues, this study aims to explore a low-cost, high-efficiency, and few-image-input method for constructing 3D point clouds of field-grown cotton, and to carry out quantitative evaluation of canopy characteristics and population structure diagnosis based on high-quality point clouds. Using UAV aerial images as the sole data source, this study innovatively integrates the Splatfacto [[28](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B28-agronomy-16-01619),[29](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B29-agronomy-16-01619)] reconstruction algorithm with the 3DGS-to-PC combined point cloud generation technique [[30](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B30-agronomy-16-01619),[31](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B31-agronomy-16-01619),[32](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B32-agronomy-16-01619)], while introducing ground control points (GCPs) placed in the field as geometric scale references. This effectively corrects scale drift during reconstruction, restores the generated cotton canopy point clouds to true physical dimensions, and ensures the accuracy and comparability of subsequent phenotypic parameter calculations. Based on this technical pathway, we systematically conduct research on the low-cost and rapid construction of 3D point clouds for field-grown cotton [[33](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B33-agronomy-16-01619)]. Compared with traditional methods, this technical pathway significantly reduces the reliance on the number of images and computational resources, achieving a technical breakthrough in reconstructing with fewer input images and high efficiency while ensuring point cloud reconstruction accuracy and completeness, greatly improving the acquisition efficiency and engineering practicality of 3D data for field cotton. On the basis of obtaining high-fidelity, true-scale 3D point clouds, we further carry out systematic quantitative evaluation of the canopy characteristics of cotton populations at the harvest stage. By extracting key canopy phenotypic indicators from point cloud data and performing quantitative calculations and comparative analyses, we systematically reveal the current structural characteristics, spatial heterogeneity patterns, and weak links of the field population canopy structure [[34](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B34-agronomy-16-01619)], and identify the core shortcomings and optimisation directions of the population structure. This provides scientific and reliable data support for precision cultivation management, canopy structure optimisation, and mechanisation-adapted planting in cotton [[35](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B35-agronomy-16-01619)]. Ultimately, the findings of this study will contribute to the sustainable development of the cotton industry towards high quality, large scale, and intelligence, providing a novel, scalable, and readily deployable technical paradigm for field crop phenotyping and population optimisation in the era of smart agriculture. Nonetheless, the present validation has been carried out only in a single season, one growth stage, and two experimental plots, and further validation involving multiple seasons and varieties is already underway to strengthen the applicability of these results.\n\n## 2. Materials and Methods\n\n### 2.1. 3D Reconstruction Workflow for Field-Grown Cotton\n\n#### 2.1.1. Data Acquisition\n\nThe data acquisition area was located at Huaxing Farm, Changji City, Changji Hui Autonomous Prefecture, Xinjiang Uygur Autonomous Region, with geographic coordinates of 44°02′ N, 87°30′ E, as shown in [Figure 1](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#fig_body_display_agronomy-16-01619-f001). The acquisition area is located in a large-scale commercial field planting zone for industrial crops. The cotton variety used in this study was Zhongmian 113GK (National Certified Cotton Variety 20253014), bred by the Institute of Cotton Research, Chinese Academy of Agricultural Sciences. This variety is a conventional cultivar of upland cotton (Gossypium hirsutum L.) with a well-defined genetic background and stable transgenic insect-resistant traits. It exhibits a relatively open plant architecture, with clear spatial distribution patterns of branching configuration and leaf area. These genotype-determined morphological characteristics directly shape the canopy spatial structure and height uniformity, providing a validation scenario with a clear biological benchmark for 3D reconstruction algorithms. Irrigation, fertilisation, and other field management practices were implemented uniformly in accordance with local standardised large-scale production protocols to ensure the reproducibility of data acquisition conditions. In this study, two plots with dimensions of 5.5 m × 5.5 m each were established, and each plot contained six standardized cotton rows with a row spacing of 0.6 m and a plant spacing of 0.1 m. The data acquisition campaign was conducted on 25 October 2025, during the cotton harvest period under clear skies, no precipitation, and stable illumination conditions, effectively ruling out the interference of meteorological variability on data quality. The average wind speed from ground level to take-off altitude was approximately 1 m\u002Fs (corresponding to Beaufort Force 1), falling within the calm to light air range. Despite the low wind speed, the cotton canopy leaves and bolls exhibited continuous slight oscillation, and since the 3D Gaussian Splatting (3DGS) reconstruction algorithm strictly relies on the static scene assumption, wind-induced deformation constituted a potential factor affecting reconstruction accuracy. The current algorithm does not explicitly model wind-induced deformation; instead, the acquisition strategy relied on two mitigation measures. First, the extremely short single exposure time (1\u002F2000 s) ensured that leaf displacement within this integration period was far smaller than the ground sampling distance, rendering motion blur within individual frames negligible. Second, the high overlap rate and short shooting intervals resulted in a baseline distance of approximately 0.5 m between adjacent along-track images, with each spatial point being covered by multiple overlapping views. The shooting interval between adjacent images was 1 s. If leaf positions exhibited observable changes due to wind during this interval, the 60% overlap rate ensured that every spatial point was recorded from a large number of views at different instants. The random positional deviations among views caused by wind were effectively averaged out through multi-constraint joint adjustment during the estimation of 3D Gaussian parameters, thereby mitigating their systematic impact on canopy structure reconstruction. A total of 141 RGB images were acquired in this campaign, all collected within a single day, ensuring the timeliness and spatial consistency of the dataset.\n\n**Figure 1.** Overview of image data acquisition for field-grown cotton. (**A**) Location map of the sampling site. (**B**) Photograph of the field experimental acquisition area. (**C**) Multi-view images of field-grown cotton acquired by the UAV.\n\n#### 2.1.2. UAV Flight Route Planning and Design\n\nIn this study, a DJI Mavic 3M UAV (DJI, Shenzhen, China) was employed for field image acquisition of field-grown cotton. The flight height was set to 5 m above ground level, and the flight speed was set to 0.5 m\u002Fs, which served as the core operational parameters for flight route planning. All acquired images were saved in JPG format, with a minimum resolution of 5280 × 3956 pixels per image. Each image was accompanied by GPS positioning, flight attitude, camera parameters, and absolute flight altitude information, providing comprehensive data support for high-precision 3D reconstruction of cotton from multiple viewpoints. A grid-based flight route was specifically designed for data acquisition in this study, and the route layout is illustrated in [Figure 2](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#fig_body_display_agronomy-16-01619-f002). Although this grid route does not possess the capability of the CCO (crop canopy-oriented) route to penetrate the occlusion of upper canopy leaves and fully capture the detailed images of middle and lower plant organs [[36](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B36-agronomy-16-01619),[37](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B37-agronomy-16-01619)], our study aimed to construct point clouds of field-grown cotton rapidly and efficiently. Therefore, the grid route allowed for the rapid acquisition of images from four viewing angles with substantially higher collection efficiency. The camera inclination angle was set to 45°, with both heading and side overlap ratios set to 60%, and the shooting interval was 1 s, thereby enabling automatic temporal continuity and precise spatial alignment.\n\n**Figure 2.** Design of the UAV-based flight route for data acquisition. (**A**) Flight trajectory of the grid route. (**B**) Distribution of image capture points along the grid route.\n\n#### 2.1.3. GCP-Based Georeferencing and Scale Calibration for 3D Canopy Reconstruction\n\nThe acquired aerial images were imported into Agisoft Metashape Professional Software (Version 2.2.2) [[38](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B38-agronomy-16-01619)] to complete the full photogrammetric processing workflow. This software, based on the Structure from Motion (SfM) algorithm, performs image stitching and registration to construct a sparse 3D point cloud model of field-grown cotton. In this study, ground control points (GCPs) were placed at the four corners of the acquisition area as georeferencing references for photogrammetric processing. The GCPs provide absolute control for the UAV images, effectively correcting geometric distortions caused by flight attitude variations, camera lens distortion, and GPS positioning errors, eliminating cumulative drift errors generated during the aerial triangulation process of 3D reconstruction, and significantly improving the absolute positioning accuracy of the model in both planar and elevational directions. More critically, the GCPs also serve a core function in establishing the true physical dimensions for the reconstructed model. During the SfM processing stage, the GCPs deployed in this study have already enabled georeferencing and absolute scale calibration of the imagery. Through bundle adjustment optimisation, a sparse point cloud model with true physical dimensions and precise camera pose information was obtained. However, the subsequent Splatfacto reconstruction model, whose input consists of normalised camera poses and image features, outputs the scene’s 3D structure in a normalised representation, the reconstructed result is constrained within a normalised coordinate space, where geometric parameters including distances and volumes are expressed in relative proportions, inherently lacking a scale correspondence to the real physical world. Consequently, although the dense point cloud produced by Splatfacto reconstruction exhibits superior local geometric details and completeness compared to the SfM sparse point cloud, its spatial scale has deviated from true physical dimensions. To address this scale loss problem, this study leverages the precisely measured GCP coordinates obtained during the SfM stage to establish a correspondence between the GCPs in the normalised point cloud space and their true physical dimensions. Specifically, based on the known physical size of the GCP markers (80 cm × 80 cm) and the distances between their corresponding corner points in the Splatfacto-reconstructed point cloud, a global scale factor between the normalised space and the real physical space is calculated. This scale factor is then applied to the entire point cloud, transforming it from normalised coordinates to true physical coordinates, thereby restoring the geometric metrics such as distances, areas, and volumes to their actual field values. Through this GCP-based post-calibration strategy, the final cotton canopy 3D point clouds not only preserve the high-precision geometric details offered by Splatfacto, but also possess a dimensionally consistent basis that matches the actual field scale, providing a reliable scale reference for subsequent accurate quantitative analyses of phenotypic parameters such as canopy height, volume fraction, and porosity [[39](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B39-agronomy-16-01619),[40](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B40-agronomy-16-01619),[41](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1619#B41-agronomy-16-01619)]. Moreover, the large 80 cm × 80 cm size significantly enhances the detectability of GCPs in UAV aerial imagery, ensuring stable and repeatable corner detection even under challenging conditions such as high flight altitudes, varying illumination, or partial canopy occlusion. This effectively reduces uncertainties in scale recovery and georeferencing, further improving the robustness and engineering practicality of the overall technical workflow. The GCP coordinates were collected using a Qianxun SR3 Pro RTK receiver (Qianxun Spatial Intelligence Inc., Shanghai, China), and the layout of the control points is shown in [Figure 3](https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F17\u002F1","MDPI Agronomy 16(17):1619 2026年8月22日","2026-08-22T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,12,9,1,"新疆农大利用UAV与3D重建技术高效评估棉花冠层，方法新颖且数据详实，对智慧农业有实质贡献。",[24],{"name":10,"url":6},[26,27,28,29,30,31],"新疆农业大学","田间棉花","3D点云","冠层性状","SfM-MVS UAV MDPI Agronomy","邱浩",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},"A6","cash-crops","棉麻糖与特色经作","\u002Ftopic\u002Ffield-crops\u002Fcash-crops","A",[],{"status":40,"locked":14,"version":41,"method":42,"evidence":43},"classified","2026-10-22.2","rules",{"candidates":44,"agricultural_evidence":47},[45,52],{"code":33,"score":46,"supported":47,"title":48,"abstract":50,"tags":51},11,true,[49],"棉花",[49],[],{"code":53,"score":54,"supported":47,"title":55,"abstract":56,"tags":58},"F5",3,[],[57],"机械化",[],[60,61],"新疆农业大学 冠层性状 田间棉花 3D点云","新疆农业大学 冠层性状","新疆农业大学冠层性状田间棉花3D点云-875","2026-08-23T00:08:13.522331Z",{"total":65,"page":21,"page_size":65,"items":66},6,[67,135,186,234,294,352],{"id":68,"title":69,"url":70,"summary":71,"summary_zh":8,"content":8,"source_name":72,"source_url":8,"published_at":73,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":74,"score_detail":75,"sources":80,"tags":82,"primary_topic":88,"secondary_topics":89,"classification":96,"search_phrases":120,"slug":123,"view_count":124,"doi":8,"paper":125,"created_at":134},5532,"精准棉花病害管理：融合无人机影像与定制YOLO深度学习框架","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.24182","研究提出单阶段全卷积检测器Cotton-Vision，融合幽灵瓶颈通道注意力与跨阶段部分融合三项轻量化设计，参数减少20%。在巴基斯坦旁遮普省7个棉田采集的2700余幅高分辨率无人机影像上测试，精确率97.7%、召回率77.7%、mAP@50为85.6%，较主流基线提升10至20个百分点。预印本，仅在巴基斯坦单一区域验证。","arXiv 预印本 2026-10-16","2026-10-16T00:00:00Z",69,{"impact":76,"substance":77,"depth":78,"authority":20,"freshness":65,"relevant":21,"comment":79},16,21,17,"方法轻量化与实测数据有增量，但仅单区域预印本验证，属细分技术进展，可入精选备选。",[81],{"name":72,"url":70},[83,84,85,86,87],"智慧农业","农业人工智能","目标检测","无人机遥感","棉花病害",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[90],{"code":91,"slug":92,"name":93,"path":94,"parent_code":95},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"status":40,"locked":14,"version":41,"method":42,"evidence":97},{"candidates":98,"agricultural_evidence":47},[99,105,110,115],{"code":91,"score":17,"supported":47,"title":100,"abstract":103,"tags":104},[101,102],"深度学习","YOLO",[],[84,85],{"code":33,"score":46,"supported":47,"title":106,"abstract":107,"tags":109},[49],[108],"cotton",[],{"code":111,"score":21,"supported":14,"title":112,"abstract":113,"tags":114},"I1",[],[],[83],{"code":116,"score":21,"supported":14,"title":117,"abstract":118,"tags":119},"I3",[],[],[86],[121,122],"Cotton-Vision 棉花病害","无人机影像 棉花 检测","Cotton-Vision棉花病害-5532",0,{"doi":8,"openalex_id":8,"authors":126,"venue":8,"cited_by_count":124,"oa_url":8,"card":127,"direction":131,"ingested_from":133},[],{"tldr":128,"method":129,"finding":130,"direction":131,"opportunity":132},"提出轻量化检测器Cotton-Vision，用无人机影像精准识别棉花病害。","基于YOLO改进，融合幽灵瓶颈通道注意力与跨阶段部分融合，用2700余幅无人机影","精确率97.7%、mAP@50达85.6%，较主流基线提升10至20个百分点。","农业遥感与作物表型","仅在巴基斯坦单一区域验证，可扩展至多区域多品种，并探索边缘设备实时部署。","agent","2026-10-22T00:07:09.816790Z",{"id":136,"title":137,"url":138,"summary":139,"summary_zh":8,"content":140,"source_name":141,"source_url":8,"published_at":142,"category":143,"cover_url":8,"hotness":13,"is_selected":14,"score":144,"score_detail":145,"sources":148,"tags":150,"primary_topic":156,"secondary_topics":157,"classification":158,"search_phrases":181,"slug":184,"view_count":124,"doi":8,"paper":8,"created_at":185},5522,"安徽农科院蚕桑所研讨「十五五」发展规划 审议全国主推技术推荐工作","https:\u002F\u002Fcss.aaas.org.cn\u002F4311487\u002F71576609.html","10月14日蚕桑所召开2026年第六次学术委员会会议，审议2027年度全国农业主导品种主推技术和重大引领性技术暨全省农业主推技术推荐工作，并专题研讨《安徽省农业科学院蚕桑研究所「十五五」发展规划》。委员们围绕家蚕种质资源保护与创制、突破性新品种选育、绿色高效种养技术与蚕桑资源综合利用等方向建言。","为充分发挥学术委员会学术评议和咨询职能，把好高层次人才引进学术关，谋划推进重大科技任务和“十五五”发展等工作，10月14日上午，安徽省农业科学院蚕桑研究所召开2026年第六次学术委员会会议。会议由所学术委员会主任李冰主持，所学术委员会全体委员参加会议。\n\n会议开展了2026年新进博士面试预答辩。新进博士依次围绕个人基本情况、学习和研究经历、代表性成果以及入职后的工作设想作PPT汇报。与会委员结合学科发展需求，从专业基础、科研能力、学科契合度、发展潜力等方面进行提问和评议，并对提出意见建议，为高层次人才引进把好学术关。\n\n会议审议2027年度全国农业主导品种主推技术和重大引领性技术暨全省农业主推技术推荐工作事宜。牵头科室就推荐工作的政策依据、推荐条件、指标分配和前期组织申报情况作说明。委员们立足蚕桑、食药用菌等学科优势，围绕拟推荐品种和技术的先进性、适用性、示范推广基础和产业带动能力进行审议，提出推荐意见，着力遴选优质成果、促进科技成果转化应用。\n\n会议专题研讨《安徽省农业科学院蚕桑研究所“十五五”发展规划》。编制组就规划编制过程、总体思路、主要目标和重点任务作说明。委员们围绕发展定位、指标设置、重点任务和保障措施提出修改建议，重点就家蚕种质资源保护与创制、突破性新品种选育、绿色高效种养技术、蚕桑资源综合利用等方向建言献策。\n\n本次会议议题聚焦、讨论充分，委员们的意见建议务实中肯，体现了学术委员会的专业水准和责任担当。会议强调，学术委员会要持续发挥专家智库作用，做强蚕桑优势学科，拓展特色经济作物研究，强化科技创新与成果服务，为安徽蚕桑和特色农业产业高质量发展提供科技支撑。","安徽省农业科学院蚕桑研究所 2026-10-19","2026-10-19T00:00:00Z","报道",60,{"impact":76,"substance":146,"depth":19,"authority":19,"freshness":65,"relevant":21,"comment":147},14,"省级科研院所学术委员会会议通稿，涉及蚕桑学科规划与全国主推技术推荐，有行业参考价值但属常规会议动态。",[149],{"name":141,"url":138},[151,152,153,154,155],"十五五规划","成果转化","种质资源","主推技术","蚕桑",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":14,"version":41,"method":42,"evidence":159},{"candidates":160,"agricultural_evidence":47},[161,165,171,176],{"code":33,"score":46,"supported":47,"title":162,"abstract":163,"tags":164},[155],[155],[],{"code":166,"score":167,"supported":47,"title":168,"abstract":169,"tags":170},"F1",5,[],[153],[154,153],{"code":172,"score":21,"supported":14,"title":173,"abstract":174,"tags":175},"G1",[],[],[152],{"code":177,"score":21,"supported":14,"title":178,"abstract":179,"tags":180},"J1",[],[],[151],[182,183],"安徽农科院 蚕桑所 十五五规划","蚕桑 主推技术 推荐","安徽农科院蚕桑所十五五规划-5522","2026-10-22T00:07:04.530022Z",{"id":187,"title":188,"url":189,"summary":190,"summary_zh":8,"content":8,"source_name":191,"source_url":8,"published_at":192,"category":143,"cover_url":8,"hotness":13,"is_selected":14,"score":144,"score_detail":193,"sources":195,"tags":197,"primary_topic":203,"secondary_topics":204,"classification":205,"search_phrases":229,"slug":232,"view_count":124,"doi":8,"paper":8,"created_at":233},5350,"中原电气实验室一行赴河南农科院烟草所考察 达成烟草烘烤节能减排合作意向","https:\u002F\u002Fhnagri.org.cn\u002Fycs\u002Farticle-114211.html","10 月 14 日,中原电气实验室能源研究所所长马志豪、电气研究所所长贾霖、顾问梁东杰等一行 5 人莅临河南农科院烟草所开展调研交流,实地参观了密集烤房设施、天敌昆虫繁育车间以及各学科实验室。双方就烟叶烘烤过程中新能源利用、储能技术应用、热能高效利用、纳米技术在农药、肥料研发等关键环节进行了深入探讨,并达成初步合作意向。","河南省农业科学院烟草研究所 2026-10-14","2026-10-14T00:00:00Z",{"impact":76,"substance":146,"depth":19,"authority":19,"freshness":65,"relevant":21,"comment":194},"省级科研机构与实验室的跨领域合作，涉及烟叶烘烤新能源与储能应用，有实质技术方向但尚处意向阶段，时效一般。",[196],{"name":191,"url":189},[198,199,200,201,202],"院企合作","节能减排","烟草烘烤","新能源利用","储能技术",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":14,"version":41,"method":42,"evidence":206},{"candidates":207,"agricultural_evidence":47},[208,213,219,225],{"code":33,"score":46,"supported":47,"title":209,"abstract":211,"tags":212},[210],"烟草",[210],[],{"code":214,"score":54,"supported":47,"title":215,"abstract":216,"tags":218},"F3",[],[217],"肥料",[],{"code":220,"score":54,"supported":47,"title":221,"abstract":222,"tags":224},"F4",[],[223],"农药",[],{"code":172,"score":21,"supported":14,"title":226,"abstract":227,"tags":228},[],[],[198],[230,231],"中原电气实验室 河南农科院烟草所","烟草烘烤 节能减排","中原电气实验室河南农科院烟草所-5350","2026-10-20T02:47:44.469410Z",{"id":235,"title":236,"url":237,"summary":238,"summary_zh":239,"content":8,"source_name":240,"source_url":237,"published_at":241,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":74,"score_detail":242,"sources":245,"tags":247,"primary_topic":250,"secondary_topics":251,"classification":252,"search_phrases":271,"slug":274,"view_count":21,"doi":275,"paper":276,"created_at":293},5234,"A smart agriculture-oriented cross-attention multiscale CNN for sugarcane disease diagnosis","https:\u002F\u002Fdoi.org\u002F10.1080\u002F23080477.2026.2749739","Sugarcane leaf diseases can adversely affect crop health and yield, making timely disease identification important for smart agriculture. Manual inspection of leaf diseases is time-consuming and requires expert knowledge, particularly when large cultivation areas need to be monitored. To tackle this problem, this study proposes a Cross-Attention Multiscale Parallel Convolutional Neural Network for automatic classification of sugarcane leaf diseases. This framework comprises a Parallel Feature Extraction Stem, cross-attention mechanism, Dense Connection Blocks, and Inception Blocks to learn disease features at various spatial scales. The PFES extracts fine and broad disease patterns using parallel convolutional branches, while the cross-attention mechanism refines the extracted features by emphasizing disease-related information. The DCB supports feature propagation and reuse, and the IB further captures multi-scale feature representations. The proposed model was tested on two benchmark datasets of leaf diseases for sugarcane plants. The proposed framework obtained an accuracy (98.79%), precision (98.31%), recall (97.81%), specificity (98.60%), and F1-Score (98.05%). For assessing the classification accuracy and consistency of the framework, the comparative study, ablation studies, and five-fold cross-validation are carried out. The results show that CA-MPCNN provides accurate sugarcane leaf disease classification and can support automated disease diagnosis in smart agriculture applications.","甘蔗叶片病害会对作物健康与产量造成不利影响，因此及时识别病害对智慧农业具有重要意义。人工检查叶片病害耗时且需要专业知识，尤其是在需要监测大面积种植区域时。为解决该问题，本研究提出一种用于甘蔗叶片病害自动分类的交叉注意力多尺度并行卷积神经网络（Cross-Attention Multiscale Parallel Convolutional Neural Network）。该框架由并行特征提取主干（Parallel Feature Extraction Stem）、交叉注意力机制、密集连接块（Dense Connection Blocks）和Inception块组成，用于学习不同空间尺度上的病害特征。PFES通过并行卷积分支提取细粒度与广域病害模式，而交叉注意力机制通过强调病害相关信息对提取的特征进行细化。DCB支持特征传播与复用，IB进一步捕获多尺度特征表示。所提出的模型在两个甘蔗植株叶片病害基准数据集上进行了测试。该框架取得了准确率（98.79%）、精确率（98.31%）、召回率（97.81%）、特异性（98.60%）和F1分数（98.05%）。为评估该框架的分类准确性和一致性，开展了对比研究、消融研究和五折交叉验证。结果表明，CA-MPCNN能够实现准确的甘蔗叶片病害分类，并可支持智慧农业应用中的自动化病害诊断。","Smart Science","2026-10-17T00:00:00Z",{"impact":19,"substance":243,"depth":76,"authority":19,"freshness":20,"relevant":21,"comment":244},20,"该研究提出一种面向甘蔗叶部病害自动分类的交叉注意力多尺度CNN，在两个基准数据集上准确率达98.79%，方法新颖、实验充分，但属于细分作物应用研究，产业影响有限，时效性强，值得进入每日精选。",[246],{"name":240,"url":237},[83,84,248,101,249],"甘蔗","病害诊断",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":14,"version":41,"method":42,"evidence":253},{"candidates":254,"agricultural_evidence":47},[255,261,266],{"code":33,"score":256,"supported":47,"title":257,"abstract":259,"tags":260},15,[258],"sugarcane",[248,258],[248],{"code":111,"score":256,"supported":47,"title":262,"abstract":264,"tags":265},[263],"smart agriculture",[83,263],[83],{"code":91,"score":267,"supported":14,"title":268,"abstract":269,"tags":270},2,[],[],[84,101],[272,273],"甘蔗 叶部病害 智能诊断","CA-MPCNN 甘蔗 病害","甘蔗叶部病害智能诊断-5234","10.1080\u002F23080477.2026.2749739",{"doi":275,"openalex_id":277,"authors":278,"venue":240,"cited_by_count":124,"oa_url":8,"card":285,"direction":291,"ingested_from":292},"W7222832245",[279,281,283],{"name":280,"orcid":8},"Bala Subramanian Chokkalingam",{"name":282,"orcid":8},"Uvarani Meenrajan",{"name":284,"orcid":8},"Shanmugapriya Sankaran",{"tldr":286,"method":287,"finding":288,"direction":289,"opportunity":290},"提出交叉注意力多尺度并行CNN，实现甘蔗叶病自动分类。","并行卷积分支、交叉注意力、密集连接与Inception块，两个基准数据集验证。","模型准确率达98.79%，消融与五折交叉验证证明其有效稳定。","农业人工智能与决策模型","可探索田间复杂背景与轻量化部署，并扩展到多作物病害实时诊断。","智慧农业 \u002F 农业物联网","openalex","2026-10-18T23:30:07.777036Z",{"id":295,"title":296,"url":297,"summary":298,"summary_zh":299,"content":8,"source_name":300,"source_url":297,"published_at":241,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":301,"score_detail":302,"sources":305,"tags":307,"primary_topic":311,"secondary_topics":312,"classification":313,"search_phrases":338,"slug":341,"view_count":21,"doi":342,"paper":343,"created_at":351},5233,"Cotton pest and disease multi-class classification with weather and temporal predictors features across Indian agroclimatic zones","https:\u002F\u002Fdoi.org\u002F10.25258\u002Fijddt.16.73s.45","The Pink Bollworm (Pectinophora gossypiella) remains one of the most destructive pests affecting cotton productivity, with infestation levels strongly driven by climatic fluctuations and crop phenology.This study presents a machine learning-based framework for early multi-classification and prediction of Pink Bollworm incidence using agro-meteorological variables.The dataset includes maximum and minimum air temperature, relative humidity, rainfall, sunshine hours, evaporation, and wind speed, all of which influence Pink Bollworm development, migration, and reproductive cycles.After preprocessing, applying cyclic encoding to temporal variables, and generating lagged climatic features, multiple machine learning algorithms-including SVM, Neural Networks, KNN, Regression models, Ensemble methods, and Naïve Bayes-were trained and systematically compared.SVM model demonstrated the best test accuracy, effectively capturing dynamic climate-pest interactions.The resulting MATLAB-based classification tool enables rapid assessment and supports climate-informed pest surveillance.A stratified 80:20 train-test split was employed, with 5-fold cross-validation on the training set for hyperparameter tuning, ensuring no data leakage between partitions.The validation and training accuracies of these models are 89.4% and 96 % respectively.Overall, the model offers a practical and interpretable solution for early detection of Pink Bollworm risk, thereby contributing to informed decision-making in integrated pest management (IPM) and precision agriculture.","棉红铃虫（Pectinophora gossypiella）仍是影响棉花产量最具破坏性的害虫之一，其发生程度受气候波动和作物物候的强烈驱动。本研究提出了一种基于机器学习的框架，利用农业气象变量对棉红铃虫的发生进行早期多分类与预测。数据集包括最高和最低气温、相对湿度、降雨量、日照时数、蒸发量和风速，这些因素均影响棉红铃虫的发育、迁移和繁殖周期。经过预处理、对时间变量应用循环编码并生成滞后气候特征后，训练并系统比较了多种机器学习算法，包括支持向量机（SVM）、神经网络、K近邻（KNN）、回归模型、集成方法和朴素贝叶斯。SVM模型表现出最佳的测试准确率，有效捕捉了气候与害虫之间的动态相互作用。所构建的基于MATLAB的分类工具可实现快速评估，并支持气候信息驱动的害虫监测。采用分层80：20的训练-测试划分，并在训练集上进行5折交叉验证以调优超参数，确保各分区之间无数据泄漏。这些模型的验证准确率和训练准确率分别为89.4%和96%。总体而言，该模型为棉红铃虫风险的早期检测提供了一种实用且可解释的解决方案，从而有助于综合害虫管理（IPM）和精准农业中的科学决策。","International Journal of Drug Delivery Technology",55,{"impact":303,"substance":17,"depth":146,"authority":65,"freshness":20,"relevant":21,"comment":304},8,"研究用多类机器学习模型结合气象与时序特征预测棉铃虫风险，方法扎实但属细分领域应用研究，信源期刊权威性一般，公共影响有限。",[306],{"name":300,"url":297},[83,308,49,309,310],"机器学习","农业气象","病虫害预警",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":14,"version":41,"method":42,"evidence":314},{"candidates":315,"agricultural_evidence":47},[316,320,327,332],{"code":33,"score":256,"supported":47,"title":317,"abstract":318,"tags":319},[108],[49,108],[49],{"code":111,"score":321,"supported":47,"title":322,"abstract":323,"tags":326},7,[],[324,325],"精准农业","precision agriculture",[83],{"code":91,"score":321,"supported":47,"title":328,"abstract":329,"tags":331},[],[308,330],"machine learning",[308],{"code":333,"score":334,"supported":47,"title":335,"abstract":336,"tags":337},"H7",4,[],[309],[309],[339,340],"棉铃虫 机器学习 预测","棉花 病虫害 气象因子","棉铃虫机器学习预测-5233","10.25258\u002Fijddt.16.73s.45",{"doi":342,"openalex_id":344,"authors":345,"venue":300,"cited_by_count":124,"oa_url":297,"card":346,"direction":291,"ingested_from":292},"W7222826491",[],{"tldr":347,"method":348,"finding":349,"direction":289,"opportunity":350},"用农业气象变量与时间特征构建机器学习模型，实现印度棉区棉铃虫多级发生预测。","气象数据加周期编码与滞后特征，比较SVM等模型，5折交叉验证。","SVM测试精度最高达89.4%，能有效刻画气候与虫害动态关系。","可引入遥感植被指数与实时气象物联网数据，构建跨区域可迁移的病虫害预警系统。","2026-10-18T23:30:07.725427Z",{"id":353,"title":354,"url":355,"summary":356,"summary_zh":357,"content":8,"source_name":358,"source_url":355,"published_at":73,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":359,"score_detail":360,"sources":363,"tags":365,"primary_topic":367,"secondary_topics":368,"classification":370,"search_phrases":387,"slug":390,"view_count":21,"doi":391,"paper":392,"created_at":426},5150,"MSFANet: a multi-scale feature aggregation network for tobacco pest recognition","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1871587","Introduction Tobacco is a globally significant economic crop, and pest infestation represents one of the primary factors contributing to declines in both yield and quality. However, tobacco pests encompass a wide variety of species that differ substantially in morphology and scale. Field-captured images are frequently subject to complex background interference and class imbalance, presenting considerable challenges for automated recognition. Existing deep learning approaches remain limited in their capacity for multi-scale feature extraction and discriminative feature representation under such conditions. To address these challenges, this study proposes a Multi-Scale Feature Aggregation Network (MSFANet) for tobacco pest recognition, formulated here as a single-label image classification task. Methods To address scale variation and complex field backgrounds, two task-oriented modules were developed. The Multi-Scale Pyramid Pooling (MSPP) module aggregates contextual information across multiple receptive fields through parallel feature extraction and adaptive cross-branch fusion. The Hierarchical Feature Fusion Block (HFFB) subsequently refines the feature representations through spatial gating and partial convolution. An auxiliary classification head and Focal Loss were incorporated during training to provide intermediate supervision and address class imbalance, respectively. Results The effectiveness of MSFANet was evaluated on a dataset comprising 13 tobacco-associated pest and natural-enemy categories. The model achieved accuracy, precision, recall, and F1-score of 97.63%, 97.89%, 97.59%, and 97.71%, respectively, with a parameter count of 1.21 M. Among the models trained from random initialization, MSFANet shared the highest accuracy while using the fewest parameters. Discussion By combining multi-scale contextual aggregation with hierarchical feature refinement, MSFANet achieved accurate classification across the 13 tobacco-associated categories under complex field conditions. The proposed method provides effective technical support for automated and precision-oriented pest monitoring in tobacco fields, offering practical value for advancing precise pest management and the broader development of smart agriculture.","引言 烟草是全球重要的经济作物，虫害侵染是导致其产量和品质下降的主要因素之一。然而，烟草害虫种类繁多，形态和尺度差异显著。田间采集的图像常受到复杂背景干扰和类别不平衡的影响，给自动化识别带来了相当大的挑战。现有深度学习方法在此类条件下进行多尺度特征提取和判别性特征表示的能力仍然有限。为应对这些挑战，本研究提出了一种用于烟草害虫识别的多尺度特征聚合网络（Multi-Scale Feature Aggregation Network，MSFANet），并将其形式化为单标签图像分类任务。方法 针对尺度变化和复杂田间背景，开发了两个面向任务的模块。多尺度金字塔池化（Multi-Scale Pyramid Pooling，MSPP）模块通过并行特征提取和自适应跨分支融合，聚合多个感受野的上下文信息。分层特征融合模块（Hierarchical Feature Fusion Block，HFFB）随后通过空间门控和部分卷积对特征表示进行细化。训练过程中引入辅助分类头和Focal Loss，分别提供中间监督和处理类别不平衡。结果 在包含13个烟草相关害虫及天敌类别的数据集上评估了MSFANet的有效性。模型的准确率、精确率、召回率和F1分数分别为97.63%、97.89%、97.59%和97.71%，参数量为1.21 M。在随机初始化训练的模型中，MSFANet以最少的参数量取得了最高准确率。讨论 通过结合多尺度上下文聚合与分层特征细化，MSFANet在复杂田间条件下实现了对13个烟草相关类别的准确分类。所提方法为烟草田间害虫的自动化和精准监测提供了有效的技术支持，对推进精准害虫管理和智慧农业的更广泛发展具有实用价值。","Frontiers in Plant Science",76,{"impact":76,"substance":77,"depth":78,"authority":361,"freshness":20,"relevant":21,"comment":362},13,"该论文提出轻量化多尺度特征聚合网络，在烟草害虫识别上取得97%以上精度且参数量仅1.21M，方法新颖、数据扎实，对智慧农业虫情监测有实用价值，值得进入每日精选。",[364],{"name":358,"url":355},[83,84,101,366,210],"病虫害识别",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[369],{"code":91,"slug":92,"name":93,"path":94,"parent_code":95},{"status":40,"locked":14,"version":41,"method":42,"evidence":371},{"candidates":372,"agricultural_evidence":47},[373,378,383],{"code":33,"score":256,"supported":47,"title":374,"abstract":376,"tags":377},[375],"tobacco",[210,375],[210],{"code":91,"score":20,"supported":47,"title":379,"abstract":380,"tags":382},[],[101,381],"deep learning",[84,101,366],{"code":111,"score":321,"supported":47,"title":384,"abstract":385,"tags":386},[],[83,263],[83],[388,389],"MSFANet 烟草害虫识别","多尺度特征聚合 害虫识别","MSFANet烟草害虫识别-5150","10.3389\u002Ffpls.2026.1871587",{"doi":391,"openalex_id":393,"authors":394,"venue":358,"cited_by_count":124,"oa_url":420,"card":421,"direction":291,"ingested_from":292},"W7222603139",[395,397,399,401,403,405,407,409,412,415,418],{"name":396,"orcid":8},"Shun He",{"name":398,"orcid":8},"Sifan Li",{"name":400,"orcid":8},"Xinjian Li",{"name":402,"orcid":8},"Qianqian Luo",{"name":404,"orcid":8},"Dong Sufen",{"name":406,"orcid":8},"Lu Lei",{"name":408,"orcid":8},"Xingjiang Chen",{"name":410,"orcid":411},"Liuti Cai","https:\u002F\u002Forcid.org\u002F0000-0002-9151-8002",{"name":413,"orcid":414},"Jian‐Yu Meng","https:\u002F\u002Forcid.org\u002F0000-0002-5796-1086",{"name":416,"orcid":417},"Ning Lu","https:\u002F\u002Forcid.org\u002F0000-0003-0214-5038",{"name":419,"orcid":8},"Hancheng Wang","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fplant-science\u002Farticles\u002F10.3389\u002Ffpls.2026.1871587\u002Fpdf",{"tldr":422,"method":423,"finding":424,"direction":289,"opportunity":425},"提出MSFANet多尺度特征聚合网络，实现复杂田间背景下烟草害虫的自动识别。","多尺度金字塔池化MSPP、分层特征融合HFFB、辅助分类头与Focal Loss","模型准确率97.63%、参数量仅1.21M，在随机初始化模型中精度最高且参数最少。","可探索轻量多尺度网络在边缘设备部署及跨作物害虫迁移识别，结合类不平衡与复杂背景的鲁棒性研究。","2026-10-17T23:30:08.695455Z"]