[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"topic-YOLOv8":3},{"name":4,"kind":5,"tokens":6,"total":7,"page":8,"page_size":9,"items":10},"YOLOv8","tag",[4],4,1,100,[11,99,159,211],{"id":12,"title":13,"url":14,"summary":15,"summary_zh":16,"content":17,"source_name":18,"source_url":14,"published_at":19,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":23,"score_detail":24,"sources":31,"tags":33,"primary_topic":17,"secondary_topics":38,"classification":39,"search_phrases":59,"slug":62,"view_count":8,"doi":63,"paper":64,"created_at":98},5136,"Enhanced YOLOv8n for adhered leaf cluster detection in RGB-based fresh green tea leaf sorting","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.atech.2026.102640","High-precision detection of adhered leaf clusters (ALCs) based on RGB imaging has long posed a critical challenge in the post-harvest sorting of fresh green tea leaves. To address this issue, this study designed an enhanced YOLOv8n model, named SDSI-YOLO, by integrating the StarNet backbone, DySample upsampling method, and Shape-Inner-IoU loss function to improve ALC detection performance. Experimental results demonstrated that SDSI-YOLO achieved an optimal balance between accuracy and efficiency, with a mean average precision of 92.6% and an F 1 -score of 89.2. It also maintained lightweight characteristics, with a model size of only 3.89 MB and efficient inference speed of 3.1 ms per frame. Ablation experiments confirmed the effectiveness of each module in constructing SDSI-YOLO. The results of visualization tests indicated that the model exhibited a stronger capability in capturing critical details and structural features of ALCs. This research provides a feasible solution for the industrial sorting of fresh green tea leaves, demonstrating considerable application potential.","基于RGB成像的高精度粘连叶簇（adhered leaf clusters，ALCs）检测一直是新鲜绿茶鲜叶采后分选中的关键难题。为解决该问题，本研究通过融合StarNet主干网络、DySample上采样方法和Shape-Inner-IoU损失函数，设计了一种增强型YOLOv8n模型，命名为SDSI-YOLO，以提升ALCs检测性能。实验结果表明，SDSI-YOLO在精度与效率之间实现了最优平衡，平均精度均值达到92.6%，F1分数为89.2。同时，该模型保持了轻量化特性，模型大小仅为3.89 MB，推理速度高达每帧3.1 ms。消融实验验证了各模块在构建SDSI-YOLO中的有效性。可视化测试结果表明，该模型在捕捉ALCs关键细节和结构特征方面表现出更强的能力。本研究为新鲜绿茶鲜叶的工业化分选提供了一种可行方案，具有相当大的应用潜力。",null,"Smart Agricultural Technology","2026-10-14T00:00:00Z","论文",10,false,76,{"impact":25,"substance":26,"depth":27,"authority":28,"freshness":29,"relevant":8,"comment":30},16,21,17,13,9,"提出轻量化改进YOLOv8n模型用于鲜茶叶粘连叶簇检测，指标与推理速度均有实测数据，方法新颖且面向产业分选场景，具备精选价值。",[32],{"name":18,"url":14},[34,35,36,4,37],"智慧农业","农业人工智能","目标检测","茶叶分选",[],{"status":40,"locked":22,"version":41,"method":42,"evidence":43},"needs_review","2026-10-22.2","rules",{"candidates":44,"agricultural_evidence":22,"reason":58},[45,53],{"code":46,"score":47,"supported":48,"title":49,"abstract":50,"tags":52},"I2",6,true,[],[51],"YOLO",[4,35,36],{"code":54,"score":8,"supported":22,"title":55,"abstract":56,"tags":57},"I1",[],[],[34],"农业相关性需要复核",[60,61],"SDSI-YOLO 茶叶分选","鲜叶 粘连叶簇 检测","SDSI-YOLO茶叶分选-5136","10.1016\u002Fj.atech.2026.102640",{"doi":63,"openalex_id":65,"authors":66,"venue":18,"cited_by_count":90,"oa_url":14,"card":91,"direction":95,"ingested_from":97},"W7221817030",[67,69,71,73,75,78,80,83,85,87],{"name":68,"orcid":17},"Jian Wang",{"name":70,"orcid":17},"Yi-jian Liu",{"name":72,"orcid":17},"Yin-yin Lin",{"name":74,"orcid":17},"Cheng-chen Zhang",{"name":76,"orcid":77},"Shengnan Liu","https:\u002F\u002Forcid.org\u002F0000-0001-9886-068X",{"name":79,"orcid":17},"Bin Ji",{"name":81,"orcid":82},"Qiu Wang","https:\u002F\u002Forcid.org\u002F0000-0002-5861-4115",{"name":84,"orcid":17},"Li-hong Du",{"name":86,"orcid":17},"Lei Zheng",{"name":88,"orcid":89},"Fei Ma","https:\u002F\u002Forcid.org\u002F0000-0002-7313-1787",0,{"tldr":92,"method":93,"finding":94,"direction":95,"opportunity":96},"提出改进YOLOv8n模型SDSI-YOLO，实现鲜茶叶粘连叶簇的高精度检测。","融合StarNet主干、DySample上采样与Shape-Inner-IoU损","mAP达92.6%、F1为89.2，模型仅3.89MB、单帧3.1ms，精度与效率兼顾。","农业人工智能与决策模型","可探索轻量检测模型在茶叶分选产线的嵌入式部署与多品类叶簇泛化能力。","openalex","2026-10-17T23:30:03.404999Z",{"id":100,"title":101,"url":102,"summary":103,"summary_zh":17,"content":104,"source_name":105,"source_url":17,"published_at":106,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":107,"score_detail":108,"sources":113,"tags":115,"primary_topic":120,"secondary_topics":126,"classification":127,"search_phrases":153,"slug":156,"view_count":157,"doi":17,"paper":17,"created_at":158},794,"方向对齐惩罚(DAP):改进YOLOv8边界框回归的轻量化定位损失(原标题:Directional Alignment Penalty: A Lightweight Localization Loss for Improved Bounding Box Regression in YOLOv8)","https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286","精准农业视觉流水线下游任务(形态特征提取、生长监测、数字孪生同步)直接依赖检测框的几何质量,因此精确定位(而非单纯正确分类)成为关键性能因素。本研究提出方向对齐惩罚(DAP)作为辅助定位正则化项,在不修改检测器架构的情况下将独立加权的归一化中心位移项引入边界框回归目标。DAP-YOLOv8使mAP@0.5:0.95从51.69%提升至53.76%、mAP@0.5从80.03%提升至80.36%,在真实世界智能温室采集的牡蛎蘑菇数据集上重复种子实验进一步表明细粒度定位的提升在不同随机初始化下一致。","## 1. Introduction\n\nSmart agriculture applications rely on computer-vision systems as the backbone for continuous visual monitoring, enabling tasks such as yield estimation, disease detection, autonomous harvesting, and environmental control. Object detection is rarely the final goal of these pipelines: the detected bounding boxes are an intermediate representation from which morphological descriptors (object size, count, maturity, and spatial distribution) are automatically derived and fused with other data streams. Thus, the quality of these descriptors is limited by the geometric accuracy of the detected boxes, and accurate localization (rather than just correct classification) becomes the key performance factor.\n\nThe YOLO family of single-stage detectors [[1](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B1-applsci-16-08286)] is widely used in this area because it predicts object locations and class probabilities in a single forward pass, providing real-time inference with competitive accuracy. In particular, YOLOv8 [[2](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B2-applsci-16-08286)] combines a decoupled detection head, a multi-scale feature aggregation neck and a hierarchical backbone, and has proven effective for continuous agricultural monitoring. The localization component of such detectors is driven by a bounding-box regression loss, and the design of this loss has a strong influence on both the convergence speed and the final localization quality.\n\nMost modern regression losses are formulated based on the intersection-over-union (IoU) between the predicted boxes and ground-truth boxes. To improve gradient behavior, IoU-based variants such as GIoU [[3](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B3-applsci-16-08286)], DIoU [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286)] and CIoU [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286),[5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B5-applsci-16-08286)] add geometric constraints based on an enclosing area, a center-distance term or an aspect-ratio penalty. However, in DIoU and CIoU, the normalized center-distance term is integrated into a single objective function with an implicit weight of one (or, for the aspect-ratio term of CIoU, with an adaptive trade-off coefficient), so it cannot be tuned, scheduled or monitored independently of the overlap and aspect-ratio terms. This limitation is most pronounced in agricultural scenes, where objects usually have irregular contours, partial occlusion and dense spatial clustering, as exemplified by the detection of oyster mushroom clusters in a cultivation chamber. To this end, we propose a Directional Alignment Penalty (DAP), an additional lightweight regularization term that explicitly encourages the predicted boxes to move toward their corresponding ground-truth boxes. DAP is measured as the Euclidean distance between the predicted and ground-truth box centers, normalized by the diagonal of the smallest box encapsulating both. We show in [Section 3.4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec3dot4-applsci-16-08286) that this quantity is, up to a square root, exactly the center-distance factor already used inside DIoU and CIoU [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286),[5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B5-applsci-16-08286)]. The contribution of DAP is to decouple that factor from the composite loss and re-express it as a separately weighted additive term whose effect on optimization can be tuned and logged independently of the base regression loss the detector already uses.\n\nThis work makes three contributions. We reformulate the normalized center-distance factor used in DIoU\u002FCIoU [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286),[5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B5-applsci-16-08286)] as an independently weighted additive regularization term that can be incorporated into existing localization losses without modifying the detector architecture. Second, we demonstrate a practical integration of the penalty into the YOLOv8 training and validation pipeline, including the separate logging of the penalty term for interpretability. Third, we evaluate the method on a dedicated oyster mushroom (Pleurotus ostreatus) dataset collected from a real-world smart greenhouse and demonstrate that stacking DAP on top of the existing CIoU-based box loss of YOLOv8 results in an additional measurable improvement in fine-grained localization, most notably a 2.07 percentage-point improvement in mAP@0.5:0.95, while preserving precision, recall, and the computational efficiency of the base detector.\n\nThe remainder of the paper is organized as follows. [Section 2](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec2-applsci-16-08286) reviews related work on single-stage detection and bounding-box regression losses. In [Section 3](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec3-applsci-16-08286), we describe the dataset, the YOLOv8 detector, the proposed DAP formulation, its formal relation to DIoU and CIoU, its integration into the detector, and the training protocol. The experimental results are presented in [Section 4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec4-applsci-16-08286). The implications and limitations of the results are discussed in [Section 5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec5-applsci-16-08286). [Section 6](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec6-applsci-16-08286) concludes the paper.\n\n## 2. Related Work\n\n### 2.1. Deep Learning-Based Object Detection\n\nDeep learning has revolutionized object detection by replacing hand-engineered feature extraction with end-to-end trainable convolutional neural networks (CNNs). Modern object detectors are commonly classified into two-stage and one-stage architectures. Two-stage detectors such as Faster R-CNN first generate region proposals and then classify and refine them, resulting in high localization accuracy but at the cost of increased computational complexity [[1](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B1-applsci-16-08286)]. In contrast, one-stage detectors perform object localization and classification in a single step, enabling substantially faster inference and making them suitable for real-time applications [[2](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B2-applsci-16-08286)].\n\nAmong one-stage detectors, the YOLO (You Only Look Once) family of detectors has become one of the most influential object detection frameworks due to its excellent speed-accuracy trade-off. The original YOLO architecture [[2](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B2-applsci-16-08286)] and its subsequent versions have continuously improved feature extraction, label assignment, multi-scale representation and localization accuracy. The most recent version, YOLOv8, is now one of the most popular object detectors in real-time computer vision applications [[3](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B3-applsci-16-08286)], thanks to its anchor-free detection approach, decoupled detection heads, improved feature fusion, and optimized training processes. However, despite these architectural improvements, the overall detection performance of YOLOv8 continues to depend strongly on the effectiveness of its bounding-box regression loss.\n\n### 2.2. IoU-Based Bounding Box Regression Losses\n\nBounding-box regression is critical to object detection because it directly determines localization accuracy. Early object detectors used coordinate-based regression losses (e.g., L1 loss and Smooth L1 loss), which considered each coordinate independently. These losses are computationally efficient but correlate weakly with the actual localization quality as measured by the Intersection over Union (IoU) metric. To overcome this limitation, Rezatofighi et al. proposed Generalized IoU (GIoU) [[3](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B3-applsci-16-08286)], which extends the conventional IoU metric by accounting for the smallest enclosing box and can be optimized even when the predicted and ground-truth boxes do not overlap. Later, Zheng et al. proposed Distance IoU (DIoU) to speed up convergence and improve localization accuracy by adding the normalized Euclidean distance between the centers of bounding boxes [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286)]. The same authors further proposed Complete IoU (CIoU) [[5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B5-applsci-16-08286)] by incorporating aspect-ratio consistency, resulting in faster convergence and better localization performance than previous IoU-based losses.\n\nBased on the success of DIoU and CIoU, many variants have been proposed to further improve bounding-box regression. Alpha-IoU [[6](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B6-applsci-16-08286)] introduces a power transformation of the IoU term to enable tunable optimization of different localization accuracies and improve robustness against noisy annotations and small datasets. MPDIoU [[7](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B7-applsci-16-08286)] replaces the classical center-distance constraint with the minimum point distances between the corresponding corners of the predicted and ground-truth boxes, which is simpler to compute and also considers the overlap, center displacement, and width-height difference [[7](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B7-applsci-16-08286)]. Among these methods, DIoU and CIoU already include a normalized center-distance term, but this term is merged into a single composite objective with an implicit, fixed weight, so its contribution to optimization is not tunable, schedulable, or monitorable independently of the overlap and aspect-ratio components. Alpha-IoU and MPDIoU jointly reweight or restructure the geometric factors, but the center-alignment contribution is still entangled with the rest of the loss. This limitation is particularly relevant for small, densely distributed, or partially occluded objects, for which the relative weight assigned to center alignment during optimization can be as important as the geometric factor itself.\n\nMotivated by this observation, we revisit the normalized center-distance term already present in DIoU and CIoU [[4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B4-applsci-16-08286),[5](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B5-applsci-16-08286)] and reformulate it as the Directional Alignment Penalty (DAP): an externally added, independently weighted, unsquared regularization component that penalizes the normalized displacement between predicted and ground-truth bounding-box centers. The contribution of DAP is not to introduce a new geometric localization signal but to decouple this factor from the composite DIoU\u002FCIoU formulation. This decoupling allows its weight to be controlled, scheduled, and logged independently of the base regression loss while maintaining full compatibility with existing IoU-based regression losses.\n\n### 2.3. Object Detection in Precision Agriculture\n\nPrecision agriculture has become one of the most promising application domains of artificial intelligence by enabling data-driven crop monitoring, disease diagnosis, yield estimation, and autonomous harvesting. The rapid development of computer vision and deep learning has significantly improved the ability of agricultural systems to detect, classify, and monitor agricultural products under real-world environmental conditions. Compared with traditional machine vision methods based on handcrafted features, deep learning-based object detectors provide greater robustness to variations in illumination, object scale, occlusion, and background complexity, making them highly suitable for agricultural applications [[8](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B8-applsci-16-08286),[9](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B9-applsci-16-08286)].\n\nRecent studies have demonstrated the successful application of deep learning-based object detection across numerous agricultural tasks, including fruit counting, crop growth monitoring, weed identification, disease detection, livestock monitoring, and greenhouse automation [[10](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B10-applsci-16-08286),[11](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B11-applsci-16-08286)]. Among the available detection frameworks, the YOLO family has become particularly attractive because of its favorable trade-off between detection accuracy and computational efficiency. Consequently, various YOLO-based architectures have been adopted for detecting apples, tomatoes, strawberries, cucumbers, peppers, grapes, citrus fruits, wheat heads, rice panicles, and other agricultural products under challenging field conditions [[12](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B12-applsci-16-08286),[13](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B13-applsci-16-08286),[14](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B14-applsci-16-08286)]. These applications require detectors capable of maintaining high localization accuracy despite illumination changes, cluttered backgrounds, overlapping objects, and varying viewing angles.\n\nOne of the most challenging agricultural detection problems is mushroom monitoring in greenhouse environments. Unlike many fruits that are relatively well separated, oyster mushrooms typically grow in dense clusters, exhibit irregular shapes, and frequently overlap at different growth stages. In addition, variations in illumination, shadows, and moisture, together with occlusion by neighboring mushrooms, further complicate accurate localization. These characteristics often reduce the effectiveness of conventional object detectors because even small localization errors may lead to inaccurate size estimation, incorrect maturity assessment, or missed detections during automated harvesting [[15](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B15-applsci-16-08286)].\n\nSeveral recent studies have explored the application of deep learning to mushroom detection and yield estimation. YOLO-based detectors have demonstrated promising performance in identifying mushroom caps, estimating maturity stages, and supporting automated harvesting systems. Researchers have also investigated improvements in feature extraction, attention mechanisms, data augmentation, and image preprocessing to increase detection accuracy under greenhouse conditions [[16](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B16-applsci-16-08286),[17](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B17-applsci-16-08286),[18](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B18-applsci-16-08286),[19](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B19-applsci-16-08286)]. Despite these advances, relatively little attention has been devoted to improving the bounding-box regression objective, even though localization quality directly influences downstream measurements such as mushroom diameter estimation, growth analysis, and digital twin-based monitoring systems. Bounding-box localization is particularly important in mushroom cultivation because multiple mushrooms frequently appear in close proximity and have partially overlapping boundaries. Under these conditions, conventional IoU-based regression losses may produce predicted bounding boxes with similar overlap scores despite differing spatial alignments with the ground truth. Such localization inconsistencies can propagate to subsequent processing stages, affecting object tracking, phenotypic measurements, and yield estimation. Therefore, improving the regression objective rather than modifying the detector architecture offers a lightweight yet effective strategy for enhancing localization accuracy while preserving real-time performance.\n\nMotivated by these difficulties, this work aims to enhance the localization capability of YOLOv8 by reinterpreting the normalized center-distance factor in DIoU\u002FCIoU as the Directional Alignment Penalty (DAP), an additional, independently weighted, unsquared regularization term rather than a novel geometric localization signal. Instead of adding network modules or increasing model complexity, DAP augments the regression loss with this decoupled and independently weighted term. Since the proposed method is only active during training and does not modify the inference pipeline, it maintains the computational efficiency of YOLOv8 while delivering more accurate bounding-box localization, as shown by the observed gains in fine-grained localization ([Section 4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec4-applsci-16-08286)). This feature makes DAP particularly well-suited for precision agriculture applications where lightweight models deployable in real time on embedded platforms are highly desirable.\n\n## 3. Materials and Methods\n\n### 3.1. Dataset\n\nThe oyster mushroom (Pleurotus ostreatus) image dataset was specifically constructed in a real-world smart mushroom greenhouse equipped with an RGB camera system and IoT-based monitoring infrastructure [[15](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B15-applsci-16-08286)]. Unlike many of the existing mushroom datasets collected under laboratory conditions or with sparse annotations, this dataset was designed to capture mushrooms in production environments, under variations in illumination, growth stage, object density, and environmental conditions. Images were acquired from two consecutive cultivation cycles in a greenhouse in Mersin, Türkiye, between late December 2022 and late April 2023 using two Full-HD (1920 × 1080) IP cameras. Images were taken during the day and at night to capture varied growth stages and lighting conditions. Each image was time-stamped to match environmental sensor readings.\n\nDuring the two cultivation cycles described above, about 1200 raw greenhouse frames were collected. After frames with no visible mushrooms and images of poor visual quality were excluded, 555 representative full-frame images were retained for object-detection experiments. These images were then split into training and validation subsets containing 499 images and 56 images, respectively. Bounding-box annotations were generated for mushroom instances visible in these full-frame images using the open-source LabelMe [[20](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B20-applsci-16-08286)] tool. Each bounding box was then refined to tightly enclose the mushroom while including minimal background. This procedure improved localization quality during supervised training. One annotation in the training set was excluded from the released label set because its class ID (2) did not match the declared two-class (immature\u002Fmature) configuration; this record was removed before training, rather than remapped, because the correct class could not be determined retrospectively.\n\nA separate, substantially larger set of approximately 8282 maturity-labeled samples was also derived from the same underlying image collection by manually cropping individual mushrooms into separate images assigned to two maturity classes, immature (5124) and mature (3158), with a balanced distribution of day (~4000) and night (~3963) instances. These cropped samples support downstream maturity-related analyses (e.g., per-instance maturity classification) and should not be interpreted as the number of full-frame images used to train the YOLO detector reported in this paper; the YOLOv8 and DAP-YOLOv8 localization experiments in [Section 4](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#sec4-applsci-16-08286) used only the 555-image, full-frame detection dataset (499 training\u002F56 validation) described above. Annotations for the full-frame detection dataset were exported in PASCAL VOC (XML), COCO (JSON), and YOLO Darknet (TXT) formats for compatibility with several detection frameworks.\n\n### 3.2. YOLOv8 Detector\n\nYOLOv8 [[2](https:\u002F\u002Fwww.mdpi.com\u002F2076-3417\u002F16\u002F16\u002F8286#B2-applsci-16-08286)] was adopted as the base detector because its single-pass design maintains high localization accuracy while greatly reducing computational cost, thereby meeting the real-time monitoring and digital-t","MDPI Applied Sciences 16(16):8286 2026-08-20","2026-08-20T00:00:00Z",71,{"impact":109,"substance":110,"depth":111,"authority":109,"freshness":29,"relevant":8,"comment":112},12,20,18,"针对农业场景提出改进YOLOv8定位损失的新方法，有具体实验支撑，但属于细分技术进展，影响范围有限。",[114],{"name":105,"url":102},[116,4,117,118,119],"精准农业","边界框回归","方向对齐惩罚","智能温室",{"code":121,"slug":122,"name":123,"path":124,"parent_code":125},"B7","protected-horticulture","设施园艺与植物工厂","\u002Ftopic\u002Fhorticulture\u002Fprotected-horticulture","B",[],{"status":128,"locked":22,"version":41,"method":42,"evidence":129},"classified",{"candidates":130,"agricultural_evidence":48},[131,137,142,149],{"code":54,"score":132,"supported":48,"title":133,"abstract":134,"tags":136},7,[],[116,135],"数字孪生",[116],{"code":121,"score":7,"supported":48,"title":138,"abstract":139,"tags":141},[],[140],"温室",[119],{"code":143,"score":144,"supported":48,"title":145,"abstract":146,"tags":148},"B5",3,[],[147],"蘑菇",[],{"code":46,"score":8,"supported":22,"title":150,"abstract":151,"tags":152},[],[],[4],[154,155],"方向对齐惩罚 边界框回归 智能温室 精准农业","方向对齐惩罚 边界框回归","方向对齐惩罚边界框回归智能温室精准农业-794",2,"2026-08-21T00:05:57.684831Z",{"id":160,"title":161,"url":162,"summary":163,"summary_zh":17,"content":164,"source_name":165,"source_url":17,"published_at":166,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":167,"score_detail":168,"sources":170,"tags":172,"primary_topic":176,"secondary_topics":181,"classification":188,"search_phrases":206,"slug":209,"view_count":144,"doi":17,"paper":17,"created_at":210},379,"TMAF-YOLO:温室番茄成熟度与缺陷果实轻量化实时检测模型","https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950","中国农业大学黄晨晓、何林然、黄文涛、张晓绚等于Sensors发表TMAF-YOLO模型——基于YOLOv8n的轻量化番茄成熟度与缺陷果实检测模型,隶属中国农业大学120周年校庆专刊。该模型以LGhostConv替换选定卷积结构减少冗余计算;引入番茄成熟度感知聚合融合模块TMAF增强检测分支的颜色、纹理和局部结构特征表示能力;使用MA-CB Focal Loss改善类别不平衡和难样本条件下的学习。实验结果显示,Precision、Recall、mAP50和mAP50-95分别达到0.889、0.873、0.954和0.776,模型仅含2.647M参数,推理速度209.030 FPS,优于YOLOv11n和YOLOv12n,同时保持实时推理性能。","## 1. Introduction\n\nTomato is an important vegetable crop in protected agriculture. Its maturity directly affects harvesting time, transport management, postharvest ripening, quality grading, and commercial value. In greenhouse environments, tomato fruits are often affected by leaf occlusion, fruit overlap, uneven illumination, and scale variation. Manual maturity assessment is inefficient and easily influenced by subjective experience. Therefore, a rapid, stable, and non-contact detection method is needed to support automated harvesting, grading management, and robotic visual perception in protected agriculture [[1](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B1-sensors-26-04950),[2](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B2-sensors-26-04950),[3](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B3-sensors-26-04950)].\n\nAmong deep learning detection methods, region-based convolutional neural network (R-CNN) models and their variants were among the earlier methods applied to fruit and vegetable detection and counting tasks. An improved Faster R-CNN model, MatDet, was proposed for multi-object tomato maturity detection in complex scenarios [[1](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B1-sensors-26-04950)]. The model used a 50-layer residual network (ResNet-50), region-of-interest alignment (RoIAlign), and a path aggregation network (PANet) to improve detection under occlusion, overlap, and illumination variation, achieving a mean average precision (mAP) of 96.14%. Faster R-CNN has also been used for tomato-related detection tasks. Highly occluded immature tomatoes on plants were detected using Faster R-CNN, with an average precision of 87.83% [[4](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B4-sensors-26-04950)]. Greenhouse tomato flowers were detected and counted using Faster R-CNN, achieving an average precision of 96.02% [[5](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B5-sensors-26-04950)]. These methods have strong localization ability and can provide stable target region information. However, their model structures are relatively complex, and their inference speed and lightweight deployment ability remain limited.\n\nWith the development of Transformer architectures, end-to-end detectors such as Detection Transformer (DETR), DETR with Improved deNoising anchOr boxes (DINO), and Real-Time Detection Transformer (RT-DETR) have been introduced into object detection tasks. DETR formulates object detection as a set prediction problem [[6](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B6-sensors-26-04950)]. DINO improves the convergence and detection performance of DETR-like models through denoising training and query optimization [[7](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B7-sensors-26-04950)]. RT-DETR further improves the Transformer detection structure for real-time detection requirements [[8](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B8-sensors-26-04950)]. In fruit and vegetable maturity recognition, PDSI-RTDETR was proposed for tomato maturity detection. Compared with the original RT-DETR, its mAP50 and mAP50-95 increased by 3.9 and 4.1 percentage points, respectively [[9](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B9-sensors-26-04950)]. An improved RT-DETR was also applied to fruit ripeness detection, indicating the potential of Transformer detectors in global feature modeling [[10](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B10-sensors-26-04950)]. However, these models usually have higher training costs, larger parameter sizes, and greater hardware requirements.\n\nIn contrast, You Only Look Once (YOLO) series models provide a better balance between detection accuracy and inference speed. They have become common methods for fruit and vegetable detection and maturity recognition. A lightweight improved YOLOv5 model was proposed for real-time tomato detection and mobile deployment [[2](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B2-sensors-26-04950)]. Compared with the original YOLOv5s, the number of parameters and floating-point operations (FLOPs) were reduced by 78% and 84.15%, respectively, while maintaining high detection accuracy. RGB-D information was combined with YOLOv5 to detect and count greenhouse tomato clusters [[3](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B3-sensors-26-04950)]. MTD-YOLO was proposed for multitask detection of cherry tomato fruits, clusters, and maturity [[11](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B11-sensors-26-04950)]. The model achieved an overall detection score of 86.6% and an average inference time of 4.9 ms. An improved YOLOX model was proposed for tomato ripeness and stem recognition, achieving an mAP of 92.17% and improving recognition under sample imbalance and scale variation [[12](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B12-sensors-26-04950)]. YOLOv8 was combined with a lightweight Swin Transformer for strawberry ripeness detection [[13](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B13-sensors-26-04950)]. Compared with YOLOv5s, CenterNet, and the Single Shot MultiBox Detector (SSD), the mAP50 increased by 1.6%, 33.5%, and 3.4%, respectively. These results indicate that improved YOLO models are suitable for fine-grained fruit maturity recognition. In addition, recent YOLO-based studies on tomato detection and ripening recognition have improved detection performance using attention mechanisms, lightweight structures, multiscale feature fusion, and postprocessing optimization [[14](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B14-sensors-26-04950),[15](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B15-sensors-26-04950),[16](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B16-sensors-26-04950),[17](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B17-sensors-26-04950)].\n\nExisting studies have verified the effectiveness of deep learning methods in fruit and vegetable detection and maturity recognition. However, further improvement is still needed for in situ greenhouse tomato detection. First, the color transition between semi-mature and mature tomatoes is continuous, making fine-grained classification difficult. Second, leaf occlusion, fruit overlap, and illumination variation weaken maturity-related feature representation. Third, defective samples are relatively limited, and conventional classification losses may underemphasize minority and hard samples. Previous studies have shown that Focal Loss and Class-Balanced Loss can adjust the loss weights of hard and imbalanced samples, thereby improving the learning ability for long-tailed classes and difficult samples [[18](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B18-sensors-26-04950),[19](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B19-sensors-26-04950)]. To address these issues, this study proposes TMAF-YOLO based on YOLOv8n.\n\nThe main contributions of this study are as follows.\n\n(1)\nA lightweight tomato maturity detection model was constructed for complex greenhouse scenarios. Inspired by the low-cost feature generation strategy in GhostNet, an LGhostConv module was designed and embedded in key positions of the Backbone and Neck [[20](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#B20-sensors-26-04950)]. This design reduces the number of parameters and computational cost while maintaining detection performance.\n\n(2)\nA tomato maturity-aware aggregation fusion module, termed TMAF, was proposed. This module was designed to address continuous color transitions between adjacent maturity stages, weak local texture differences, and insufficient feature representation caused by occlusion and overlap. By integrating local texture enhancement and channel-spatial recalibration, TMAF strengthens maturity-related feature representation.\n\n(3)\nMA-CB Focal Loss was developed to improve learning under class imbalance and hard-sample conditions. By combining class-balanced weights with the Focal Loss modulation term, the loss enhances learning for minority and difficult samples, including defective fruits and adjacent maturity stages.\n\n## 2. Materials and Methods\n\n### 2.1. Overall Method Design\n\nThis study focuses on in situ tomato maturity detection before greenhouse harvesting. An overall application workflow was constructed, including greenhouse image acquisition, fruit detection and category recognition, harvesting decision-making, and postharvest grading management. As shown in [Figure 1](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#fig_body_display_sensors-26-04950-f001), tomato images were first acquired under natural growth conditions in the greenhouse. Then, the developed detection model was used to identify fruit locations, maturity stages, and defective fruits. Finally, the detection results were used to support harvesting time determination, defective fruit screening, and postharvest grading and quality management. It should be noted that this study mainly focuses on greenhouse tomato maturity and defective fruit detection. The harvesting, grading, and management processes in the figure are used to illustrate the potential application scenarios of the proposed model.\n\n**Figure 1.** Overall workflow of greenhouse tomato maturity detection.\n\n### 2.2. Dataset Construction and Annotation\n\nThe self-collected images used in this study were acquired in a tomato greenhouse at the Tianjin Academy of Agricultural Sciences, China, from 10 to 17 January 2026. Image acquisition was conducted between 9:00 a.m. and 12:00 p.m. using a HUAWEI Mate 60 smartphone (Huawei Device Co., Ltd., Dongguan, China) in automatic shooting mode. Exposure, focus, ISO sensitivity, and shutter speed were adjusted automatically by the built-in camera system. The shooting distance ranged from approximately 0.5 to 2.0 m, and the image resolution was 1280 × 760 pixels. Images were captured under naturally varying greenhouse illumination; illuminance was not measured instrumentally.\n\nTo represent practical greenhouse conditions, images were acquired from different viewing angles and distances. The collected scenes included fruits at different maturity stages and scales, together with leaf and branch occlusion, fruit overlap, dense fruit distribution, small targets, complex backgrounds, and local image blur. Representative examples of self-collected images are shown in [Figure 2](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#fig_body_display_sensors-26-04950-f002).\n\n**Figure 2.** Representative tomato images acquired under typical greenhouse conditions: (**a**) mature fruits; (**b**) adjacent fruits at different maturity stages; (**c**) fruit overlap and foliage interference; (**d**) leaf and branch occlusion; (**e**) small targets in a relatively long-range view; and (**f**) local image blur.\n\nA total of 596 valid images were collected from the greenhouse. Because defective fruits were relatively scarce, 33 additional defective-fruit images were selected from the Tomato Fruit Diseases Dataset available on Kaggle ([https:\u002F\u002Fwww.kaggle.com\u002Fdatasets\u002Fprofnourasemary\u002Ftomato-dataset](https:\u002F\u002Fwww.kaggle.com\u002Fdatasets\u002Fprofnourasemary\u002Ftomato-dataset), accessed on 1 July 2026). The additional images mainly consisted of tomato fruits exhibiting visible disease symptoms, together with a smaller number showing physical damage, decay, cracking, or other abnormal surface characteristics. The selected public images were annotated according to the same criteria as the self-collected images.\n\nThe 596 self-collected images were divided at the original-image level into training, validation, and test subsets containing 470, 63, and 63 images, respectively. The 33 public images were added exclusively to the training set to increase the number and diversity of defective-fruit samples. Consequently, the final training, validation, and test sets contained 503, 63, and 63 images, respectively. The validation and test sets therefore consisted entirely of self-collected greenhouse images.\n\nThe annotated tomato instances were divided into four categories: immature, semi-mature, mature, and defective. The first three categories represent successive normal maturity stages based primarily on fruit color and visible appearance. The defective category includes fruits with visible disease spots, physical damage, decay, cracking, or other abnormal surface characteristics and was treated as an independent appearance category rather than as a normal maturity stage.\n\nAll tomato instances were manually annotated with bounding boxes using LabelMe v6.0.0. The original annotations were stored in JSON format and subsequently converted into the format required by the YOLO detection framework. Partially occluded or locally blurred fruits were retained when their categories and principal visible boundaries could be reliably determined. Instances were excluded only when severe occlusion, insufficient visible information, or ambiguous appearance prevented consistent annotation. All annotations were reviewed according to uniform class definitions to ensure labeling consistency.\n\nThe final dataset contained 629 images and 2505 annotated tomato instances, including 1170 immature, 788 semi-mature, 304 mature, and 243 defective instances. Data augmentation was performed after dataset partitioning and was applied only to the training set. The validation and test sets were not augmented. The training set was used for model parameter optimization, the validation set for training monitoring and model selection, and the test set exclusively for final performance evaluation. The image and instance distributions before data augmentation are summarized in [Table 1](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#table_body_display_sensors-26-04950-t001).\n\n**Table 1.** Distribution of images and annotated tomato instances before data augmentation.\n\n### 2.3. Construction of the TMAF-YOLO Model\n\n#### 2.3.1. TMAF-YOLO: An Improved YOLOv8n Model\n\nThe proposed TMAF-YOLO uses YOLOv8n as the baseline model. It is improved from three aspects: lightweight feature extraction, maturity-related feature fusion, and classification loss optimization. The overall architecture is shown in [Figure 3](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#fig_body_display_sensors-26-04950-f003). After the original image is input into the network, multiscale features are first extracted by the Backbone, in which selected convolutional structures are replaced by LGhostConv. Then, features at different scales are fused in the Neck, and TMAF modules are added to the P3\u002F8 and P4\u002F16 branches. Finally, the model outputs prediction results through three detection heads at P3\u002F8, P4\u002F16, and P5\u002F32 scales. During training, the model is optimized using a joint loss function composed of WIoU, DFL, and MA-CB Focal Loss. The specific designs of LGhostConv, TMAF, and MA-CB Focal Loss are introduced in the following sections.\n\n**Figure 3.** Overall architecture of TMAF-YOLO.\n\n#### 2.3.2. Lightweight Feature Extraction Module: LGhostConv\n\nTo reduce redundant computation during feature extraction and multiscale feature fusion, a lightweight Ghost convolution module, termed LGhostConv, was introduced into selected downsampling layers of the backbone and neck. Similar to GhostConv, LGhostConv generates output features through primary and complementary feature branches. However, it replaces the dense spatial convolution in the primary branch with a depthwise-separable operation and employs directional depthwise convolutions in the complementary branch.\n\nGiven an input feature map $X$ with $C_{i n}$ channels, LGhostConv first applies a $k \\times k$ depthwise convolution with stride s for spatial feature extraction and downsampling, followed by a $1 \\times 1$ pointwise convolution for channel interaction. This process generates approximately half of the required output channels as the primary features. The remaining channels are generated by dividing the primary features into two groups and applying $1 \\times 3$ and $3 \\times 1$ depthwise convolutions to capture horizontal and vertical information, respectively. The directional features are then fused with the primary features through a parameter-free residual connection and concatenated to form the final output.\n\nIgnoring bias terms, batch-normalization parameters, and activation operations, the number of parameters in LGhostConv is expressed as follows:\n\n$$\nP_{L G h o s t C o n v} = k^{2} C_{i n} + C_{i n} m + 3 n\n$$\n\n(1)\n\nwhere $m$ = ⌈$C_{o u t}$\u002F2⌉ and $n$ = $C_{o u t}$ − $m$ denote the numbers of primary and complementary feature channels, respectively. In comparison, a standard $k \\times k$ convolution requires $k^{2} C_{i n} C_{o u t}$ parameters. Therefore, LGhostConv reduces model complexity by replacing dense spatial channel interactions with depthwise spatial filtering, pointwise channel projection, and low-cost directional feature generation. Its effects on detection accuracy, model size, computational complexity, and inference speed are evaluated in [Section 3.3](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#sec3dot3-sensors-26-04950).\n\n#### 2.3.3. Tomato Maturity-Aware Aggregation Fusion Module: TMAF\n\nTo strengthen the representation of maturity- and defect-related cues in the fused neck features, a tomato maturity-aware aggregation fusion module, termed TMAF, was designed. TMAF is embedded in the P3 and P4 feature fusion paths of the YOLOv8n neck, where relatively high-resolution features are retained for fine-grained maturity and defect detection. The original P5 path is preserved to limit the additional computational cost. As illustrated in [Figure 4](https:\u002F\u002Fwww.mdpi.com\u002F1424-8220\u002F26\u002F15\u002F4950#fig_body_display_sensors-26-04950-f004), TMAF comprises a lightweight texture aggregation branch and a TMCAttention submodule.\n\n**Figure 4.** Structure of the TMAF module.\n\nLet the input feature be $X$, with a size of $X \\in R^{B \\times C \\times H \\times W}$, where $B$, $C$, $H$, and $W$ denote the batch size, number of channels, feature map height, and feature map width, respectively. The input feature first passes through a $1 \\times 1$ convolution for channel mapping. It is then split along the channel dimension into a retained branch and a texture branch:\n\n$$\nX_{1} , X_{2} = S p l i t \\left(\\right. C o n v_{1 \\times 1} \\left(\\right. X \\left.\\right) \\left.\\right)\n$$\n\n(2)\n\nwhere $X_{1}$ is the retained branch, which maintains the direct transmission of input information. $X_{2}$ is the texture branch, which passes through three LTMBlocks in sequence to obtain texture-enhanced features at different levels, denoted as $T_{1}$, $T_{2}$, and $T_{3}$. Each LTMBlock consists of a $1 \\times 1$ convolution, a $3 \\times 3$ depthwise convolution, a $5 \\times 5$ depthwise convolution, and a $1 \\times 1$ convolution. The depthwise convolutions are used to supplement local texture and edge details. The $1 \\times 1$ convolutions are used for channel adjustment and feature fusion.\n\nThen, the retained branch $X_{1}$, and the texture branch outputs $T_{1}$, $T_{2}$, and $T_{3}$ are concatenated along the channel dimension. A $1 \\times 1$ convolution is then used to complete feature fusion and obtain the fused feature $F$:\n\n$$\n= C o n v_{1 \\times 1} \\left(\\right. C o n c a t \\left(\\right. X_{1} , T_{1} , T_{2} , T_{3} \\left.\\right) \\left.\\right)\n$$\n\n(3)\n\nTo further adjust maturity-related feature responses, TMCAttention is applied after feature fusion. This structure contains three branches: channel attention, spatial attention, and residual connection. Channel attention is used to adjust the importance of different feature channels. It helps the model focus on color, texture, and defect responses related to maturity discrimination. Spatial attention is used to highlight fruit region responses and reduce interference from irrelevant background features. The outputs of the two attention branches are denoted as $F_{c}$ and $F_{s}$, respectively.\n\nFinally, the outputs of the channel attention branch and the spatial attention branch are fused and added to the residual branch. The output feature of the TMAF module is obtained as follows:\n\n$$\nY = F + C o n v_{1 \\times 1} \\left(\\right. F_{c} + F_{s} \\left.\\righ","Sensors 2026-08-05","2026-08-05T00:00:00Z",62,{"impact":109,"substance":111,"depth":25,"authority":109,"freshness":7,"relevant":8,"comment":169},"针对温室番茄成熟度与缺陷检测的轻量化模型研究，方法有创新，但影响范围限于细分领域，时效性一般。",[171],{"name":165,"url":162},[173,174,4,175],"中国农业大学","Sensors","番茄成熟度",{"code":177,"slug":178,"name":179,"path":180,"parent_code":125},"B1","vegetables","蔬菜","\u002Ftopic\u002Fhorticulture\u002Fvegetables",[182,187],{"code":46,"slug":183,"name":184,"path":185,"parent_code":186},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"code":121,"slug":122,"name":123,"path":124,"parent_code":125},{"status":128,"locked":22,"version":41,"method":42,"evidence":189},{"candidates":190,"agricultural_evidence":48},[191,195,201],{"code":46,"score":109,"supported":48,"title":192,"abstract":193,"tags":194},[51],[51],[4],{"code":177,"score":196,"supported":48,"title":197,"abstract":199,"tags":200},11,[198],"番茄",[198],[],{"code":121,"score":202,"supported":48,"title":203,"abstract":204,"tags":205},8,[140],[],[],[207,208],"中国农业大学 番茄成熟度 Sensors YOLOv8","中国农业大学 番茄成熟度","中国农业大学番茄成熟度SensorsYOLOv8-379","2026-08-10T23:54:46.339395Z",{"id":212,"title":213,"url":214,"summary":215,"summary_zh":17,"content":216,"source_name":217,"source_url":17,"published_at":166,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":218,"score_detail":219,"sources":223,"tags":225,"primary_topic":17,"secondary_topics":230,"classification":231,"search_phrases":256,"slug":259,"view_count":144,"doi":17,"paper":17,"created_at":260},374,"轻量化GSP-YOLO实时杂草识别模型——激光除草场景97.74%精确率、41 FPS推理速度","https:\u002F\u002Fagritechinsights.com\u002Findex.php\u002F2026\u002F08\u002F05\u002Fai-model-gsp-yolo-slashes-weed-battles-without-harming-crops","南京市农业机械化研究所刘赛赛等在Smart Agricultural Technology发表GSP-YOLO模型——基于YOLOv8s的轻量化杂草实时检测系统。该模型将YOLOv8s标准骨干替换为GhostNetV3,并集成空间-通道注意力模块(SCSA),提升对小型、形态相似杂草幼苗的识别能力。在NIAM9weeds数据集上达97.74%精确率,优于YOLOv5n和YOLOv10s等轻量化模型,模型大小减少近30%,计算需求减少超过30%。在边缘设备上通过TensorRT加速,处理速度41 FPS,可支持以步行速度移动的激光除草机器人。温室试验中,系统在0.72 km\u002Fh速度下达93.4%杂草识别准确率,平均275毫秒\u002F株完成除草,作物损伤率低于3.4%。","Researchers in Nanjing have developed a lightweight AI model that could revolutionize how farmers tackle weeds—without harming crops. The breakthrough, published in Smart Agricultural Technology and led by Liu Saisai of the Nanjing Institute of Agricultural Mechanization, introduces GSP-YOLO, a streamlined version of the popular YOLO object detection system, tailored for real-time weed detection in dynamic field conditions.\n\nWeeds may be small, but they’re persistent—and costly. Traditional weeding methods often rely on herbicides or broad-spectrum mechanical removal, which can damage surrounding crops and contribute to environmental strain. Laser weeding, a targeted alternative, requires precise, real-time identification of weed seedlings—especially challenging when weeds look nearly identical to young crops and must be distinguished in seconds.\n\nGSP-YOLO steps in with a targeted solution. By replacing the standard backbone of YOLOv8s with GhostNetV3—a network designed for efficiency—and integrating a new spatial-channel attention module (SCSA), the model sharpens its ability to spot tiny, morphologically similar weed seedlings. “The challenge isn’t just detecting weeds,” says Liu. “It’s doing it fast, accurately, and on a device small enough to ride on a tractor or robot—without draining battery or computing power.”\n\nThe numbers back up the claim. On the NIAM9weeds dataset, GSP-YOLO hit 97.74% precision, outperforming other lightweight models like YOLOv5n and YOLOv10s. Even more striking: it did so while shrinking the model size by nearly 30% and cutting computational demands by over 30%. When deployed on edge devices and accelerated with TensorRT, it processed images at 41 frames per second—fast enough for a laser weeding robot moving at walking speed.\n\nIn live greenhouse trials, the system achieved 93.4% weed recognition accuracy at 0.72 km\u002Fh, removing weeds in an average of 275 milliseconds each, with crop injury kept below 3.4%. That speed and precision could make laser weeding commercially viable for the first time outside small-scale or research settings.\n\nFor farmers, the implications are significant. Laser weeding promises chemical-free, site-specific control that preserves soil health and biodiversity. But until now, the technology has been held back by bulky, slow, or inaccurate detection systems. GSP-YOLO changes that equation. By making AI-powered, real-time weed detection both accurate and deployable on low-power edge devices, it opens the door to scalable, autonomous weeding systems that could operate across thousands of acres.\n\nIndustry observers see this as a turning point. “What we’re seeing here is not just an incremental improvement,” notes one agritech analyst. “It’s a fundamental shift in how we can integrate AI into field operations—lightweight enough for the farm, powerful enough for the task.”\n\nAs robotic platforms become more common in agriculture, models like GSP-YOLO could become the “eyes” of autonomous weeding systems, enabling precise, eco-friendly control in row crops, orchards, and greenhouses. Future developments might see this technology adapted for drone-based scouting or integrated into variable-rate spraying systems, further reducing chemical use.\n\nFor now, Liu and the team are focused on real-world deployment. “Our goal isn’t just to detect weeds,” Liu explains. “It’s to make sure every farmer, regardless of scale, can use this technology to grow more with less—cleaner, smarter, and more sustainably.”\n\nWith precision agriculture evolving rapidly, innovations like GSP-YOLO remind us that sometimes, the smallest models can have the biggest impact.","Smart Agricultural Technology 2026-08-05",77,{"impact":110,"substance":220,"depth":111,"authority":109,"freshness":221,"relevant":8,"comment":222},22,5,"轻量化AI模型在激光除草领域取得突破，性能数据详实，对精准农业有实质推动，但时效性稍差。",[224],{"name":217,"url":214},[226,4,227,228,229],"南京农机化所","激光除草","边缘部署","SmartAgTech",[],{"status":40,"locked":22,"version":41,"method":42,"evidence":232},{"candidates":233,"agricultural_evidence":48,"reason":255},[234,240,244,251],{"code":235,"score":109,"supported":48,"title":236,"abstract":238,"tags":239},"F4",[237],"除草",[237],[227],{"code":46,"score":109,"supported":48,"title":241,"abstract":242,"tags":243},[51],[51],[4],{"code":245,"score":47,"supported":48,"title":246,"abstract":247,"tags":250},"F5",[],[248,249],"农业机械","机械化",[],{"code":121,"score":144,"supported":48,"title":252,"abstract":253,"tags":254},[],[140],[],"多个产业证据相当，需要确认主归属",[257,258],"南京农机化所 激光除草 边缘部署 SmartAgTech","南京农机化所 激光除草","南京农机化所激光除草边缘部署SmartAgTech-374","2026-08-10T23:54:46.220254Z"]