[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-379":3,"related-379":78},{"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":30,"secondary_topics":36,"classification":48,"search_phrases":72,"slug":75,"view_count":76,"doi":8,"paper":8,"created_at":77},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。",null,"## 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","论文",10,false,62,{"impact":17,"substance":18,"depth":19,"authority":17,"freshness":20,"relevant":21,"comment":22},12,18,16,4,1,"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。",[24],{"name":10,"url":6},[26,27,28,29],"中国农业大学","Sensors","YOLOv8","番茄成熟度",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},"B1","vegetables","蔬菜","\u002Ftopic\u002Fhorticulture\u002Fvegetables","B",[37,42],{"code":38,"slug":39,"name":40,"path":41,"parent_code":35},"B7","protected-horticulture","设施园艺与植物工厂","\u002Ftopic\u002Fhorticulture\u002Fprotected-horticulture",{"code":43,"slug":44,"name":45,"path":46,"parent_code":47},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"status":49,"locked":14,"version":50,"method":51,"evidence":52},"classified","2026-10-22.2","rules",{"candidates":53,"agricultural_evidence":55},[54,60,66],{"code":43,"score":17,"supported":55,"title":56,"abstract":58,"tags":59},true,[57],"YOLO",[57],[28],{"code":31,"score":61,"supported":55,"title":62,"abstract":64,"tags":65},11,[63],"番茄",[63],[],{"code":38,"score":67,"supported":55,"title":68,"abstract":70,"tags":71},8,[69],"温室",[],[],[73,74],"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",3,"2026-08-10T23:54:46.339395Z",{"total":79,"page":21,"page_size":79,"items":80},6,[81,136,200,294,370,431],{"id":82,"title":83,"url":84,"summary":85,"summary_zh":8,"content":8,"source_name":86,"source_url":8,"published_at":87,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":88,"score_detail":89,"sources":94,"tags":96,"primary_topic":102,"secondary_topics":103,"classification":105,"search_phrases":121,"slug":124,"view_count":125,"doi":8,"paper":126,"created_at":135},5536,"日光温室番茄全生育期环境特征解析与智能调控决策支持方法","https:\u002F\u002Fpaper.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F20261021145441802156759.shtm","研究解析日光温室番茄全生育期环境特征，发现该类温室气候缓冲能力较强、有效积温与番茄生理需求呈双峰匹配，但相对增益指标揭示单位时间热量供给失衡、深冬热量不足以及光能结构性不足、湿度调控昼夜失衡等问题。作者建议构建温度—光照—湿度耦合的智能闭环控制系统。","Agriculture 2026-10-21","2026-10-21T00:00:00Z",77,{"impact":19,"substance":90,"depth":91,"authority":92,"freshness":13,"relevant":21,"comment":93},21,17,13,"论文解析日光温室番茄全生育期环境特征并提出温光湿耦合智能调控方案，结论有新意，对设施农业信息化有参考价值。",[95],{"name":86,"url":84},[97,98,99,100,101],"智慧农业","设施农业","环境调控","温室番茄","智能决策",{"code":38,"slug":39,"name":40,"path":41,"parent_code":35},[104],{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},{"status":49,"locked":14,"version":50,"method":51,"evidence":106},{"candidates":107,"agricultural_evidence":55},[108,112,116],{"code":38,"score":92,"supported":55,"title":109,"abstract":110,"tags":111},[69],[69],[100,99],{"code":31,"score":61,"supported":55,"title":113,"abstract":114,"tags":115},[63],[63],[],{"code":117,"score":21,"supported":14,"title":118,"abstract":119,"tags":120},"I1",[],[],[97],[122,123],"日光温室 番茄 环境调控","温室番茄 智能闭环控制","日光温室番茄环境调控-5536",0,{"doi":8,"openalex_id":8,"authors":127,"venue":8,"cited_by_count":125,"oa_url":8,"card":128,"direction":132,"ingested_from":134},[],{"tldr":129,"method":130,"finding":131,"direction":132,"opportunity":133},"解析日光温室番茄全生育期环境特征并提出智能调控决策支持方法。","基于全生育期环境监测数据，分析积温匹配与相对增益指标。","温室缓冲力强但深冬热量不足、光能结构性不足、湿度昼夜失衡。","智慧农业 \u002F 农业物联网","可研发温度—光照—湿度耦合的智能闭环控制算法，并验证其对番茄产量与能耗的协同优化效果。","agent","2026-10-22T00:07:10.129353Z",{"id":137,"title":138,"url":139,"summary":140,"summary_zh":141,"content":8,"source_name":142,"source_url":139,"published_at":143,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":144,"score_detail":145,"sources":147,"tags":149,"primary_topic":154,"secondary_topics":155,"classification":157,"search_phrases":178,"slug":181,"view_count":125,"doi":182,"paper":183,"created_at":199},5387,"LIGHTWEIGHT FEATURE FUSED ATTENTION DRIVEN MULTI LEVEL HYBRID MOBILENET V2 FRAMEWORK FOR EARLY TOMATO LEAF DISEASE DETECTION","https:\u002F\u002Fdoi.org\u002F10.62441\u002Factainnovations.vi.1918","It is crucial to find and treat tomato crop diseases early and accurately to improve crop efficiency and make sure there is no scarcity of crop. However, computerized analysis is provocative due to lighting differences, compound backgrounds, and graphic likenesses among infection regions. Research work proposed a lightweight Attention Driven Multi Level Hybrid MobileNet-V2 algorithm that fuses the handcrafted texture structures with Deep Learning (DL) features, then image preprocessing is incorporated using Gaussian, Bilateral, and Guided filters to minimize noise by preserving the disease lesions. LAB color model along with Contrast Limited Adaptive Histogram Equalization (CLAHE) are used to improve the contrast. Otsu thresholding is involved to effectively separate the leaf segment from diseased parts. Local Binary Pattern (LBP) and Gray Level Co-occurrence Matrix (GLCM) descriptors are used to obtain fine-grained features, while high level features are gained by pretrained MobileNet-V2 model. Features fusion is fed to the fully connected layers of deep learning model for training and validation purpose. Results have revealed that the suggested model attains better classification performance as compared to existing models, making it an effective explanation for real time tomato leaf disease detection for smart agriculture applications.","早期准确发现并治疗番茄作物病害，对于提高作物效率、确保作物不短缺至关重要。然而，由于光照差异、复杂背景以及感染区域之间的图形相似性，计算机化分析具有挑战性。研究提出了一种轻量级注意力驱动多层混合MobileNet-V2算法，该算法将手工纹理结构与深度学习（DL）特征相融合，然后结合高斯、双边和引导滤波器进行图像预处理，以在保留病害病变的同时最小化噪声。采用LAB颜色模型和对比度受限自适应直方图均衡化（CLAHE）来提高对比度。引入Otsu阈值分割以有效分离叶片片段与病害部分。使用局部二值模式（LBP）和灰度共生矩阵（GLCM）描述符获取细粒度特征，同时通过预训练的MobileNet-V2模型获取高层特征。特征融合后输入深度学习模型的全连接层进行训练和验证。结果表明，与现有模型相比，所提出的模型取得了更好的分类性能，使其成为智慧农业应用中实时番茄叶片病害检测的有效解决方案。","Acta Innovations","2026-10-18T00:00:00Z",64,{"impact":17,"substance":18,"depth":19,"authority":13,"freshness":67,"relevant":21,"comment":146},"提出融合手工纹理特征与MobileNet-V2的轻量番茄叶病检测方法，有一定方法新颖性，但属单篇论文、应用范围有限，未达每日精选门槛。",[148],{"name":142,"url":139},[97,150,151,152,153],"农业人工智能","深度学习","番茄病害","图像识别",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[156],{"code":43,"slug":44,"name":45,"path":46,"parent_code":47},{"status":49,"locked":14,"version":50,"method":51,"evidence":158},{"candidates":159,"agricultural_evidence":55},[160,166,172],{"code":31,"score":161,"supported":55,"title":162,"abstract":164,"tags":165},15,[163],"tomato",[63,163],[152],{"code":43,"score":167,"supported":55,"title":168,"abstract":169,"tags":171},9,[],[151,170],"deep learning",[150,153,151],{"code":117,"score":173,"supported":55,"title":174,"abstract":175,"tags":177},7,[],[97,176],"smart agriculture",[97],[179,180],"MobileNet-V2 番茄叶片病害","番茄叶病 深度学习 检测","MobileNet-V2番茄叶片病害-5387","10.62441\u002Factainnovations.vi.1918",{"doi":182,"openalex_id":184,"authors":185,"venue":142,"cited_by_count":125,"oa_url":191,"card":192,"direction":132,"ingested_from":198},"W7222977034",[186,188],{"name":187,"orcid":8},"Mrs. Vijetalaxmi A Hegde",{"name":189,"orcid":190},"Abhay Ramakant Kasetwar","https:\u002F\u002Forcid.org\u002F0000-0002-5800-8764","https:\u002F\u002Factainnovations.com\u002Findex.php\u002Fpub\u002Farticle\u002Fdownload\u002F1918\u002F981",{"tldr":193,"method":194,"finding":195,"direction":196,"opportunity":197},"提出轻量注意力多级混合MobileNet-V2框架，融合手工纹理与深度学习特征实现番茄叶片病害早期检","高斯\u002F双边\u002F引导滤波去噪，LAB+CLAHE增强，Otsu分割，LBP与GLCM","所提模型分类性能优于现有模型，可有效用于智慧农业实时番茄叶片病害检测。","农业人工智能与决策模型","可探索多作物跨域泛化、田间复杂光照下轻量模型部署与可解释性，及病害早期分级检测。","openalex","2026-10-20T23:30:10.424511Z",{"id":201,"title":202,"url":203,"summary":204,"summary_zh":205,"content":8,"source_name":206,"source_url":203,"published_at":207,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":208,"score_detail":209,"sources":211,"tags":213,"primary_topic":217,"secondary_topics":218,"classification":225,"search_phrases":271,"slug":274,"view_count":125,"doi":8,"paper":275,"created_at":293},5313,"Real-Time IoT-Integrated Ensemble LSTM–Bi-LSTM Framework for Greenhouse Microclimate Forecasting","https:\u002F\u002Fwasrpublication.com\u002Findex.php\u002Findex\u002Farticle\u002Fview\u002F512","Accurate prediction of greenhouse microclimates is essential for optimizing crop growth and resource utilization in precision agriculture. The research presents an IoT-integrated ensemble deep learning framework for real-time greenhouse microclimate forecasting using Long Short-Term Memory (LSTM) and Bidirectional Long Short-Term Memory (Bi-LSTM) networks. Environmental variables, including temperature, light intensity, soil moisture, and humidity, were acquired through IoT sensor nodes deployed in a controlled greenhouse environment. A benchmark greenhouse dataset was first used for model development and hyperparameter optimization, while an original IoT-based dataset collected during cucumber, tomato, and pepper cultivation at the Federal University Wukari, Nigeria, was used for validation and real-world deployment. The data were preprocessed through noise filtering, missing value imputation, Z-score-based outlier removal, and Min–Max normalization. A weighted ensemble strategy was then applied to combine LSTM and Bi-LSTM predictions for improved robustness. Experimental results show that the Bi-LSTM model outperformed the LSTM model, achieving R² = 0.9981, MAE = 0.0088, RMSE = 0.0110, and MSE = 0.0001, compared to LSTM results of R² = 0.9956, MAE = 0.0133, RMSE = 0.0167, and MSE = 0.0003. Statistical validation using paired t-test, Wilcoxon signed-rank test, confidence intervals, and multi-run analysis confirmed that the performance improvement is statistically significant and robust. The proposed framework enables accurate real-time microclimate forecasting and intelligent environmental monitoring, providing a scalable solution for precision greenhouse management, improved crop productivity, optimized energy utilization, and efficient irrigation scheduling.","准确预测温室微气候对于精准农业中优化作物生长和资源利用至关重要。本研究提出了一种集成物联网的集成深度学习框架，利用长短期记忆网络（LSTM）和双向长短期记忆网络（Bi-LSTM）进行实时温室微气候预测。温度、光照强度、土壤湿度和湿度等环境变量通过部署在受控温室环境中的物联网传感器节点采集。研究首先使用基准温室数据集进行模型开发和超参数优化，随后使用在尼日利亚武卡里联邦大学黄瓜、番茄和辣椒种植期间采集的原始物联网数据集进行验证和实际部署。数据经过噪声滤波、缺失值插补、基于Z分数的异常值去除和最小—最大归一化预处理。然后采用加权集成策略融合LSTM和Bi-LSTM的预测结果以提高鲁棒性。实验结果表明，Bi-LSTM模型优于LSTM模型，其R² = 0.9981、MAE = 0.0088、RMSE = 0.0110、MSE = 0.0001，而LSTM的结果为R² = 0.9956、MAE = 0.0133、RMSE = 0.0167、MSE = 0.0003。采用配对t检验、Wilcoxon符号秩检验、置信区间和多轮运行分析进行的统计验证证实，性能提升具有统计显著性和鲁棒性。所提出的框架能够实现准确的实时微气候预测和智能环境监测，为精准温室管理、提高作物生产力、优化能源利用和高效灌溉调度提供了可扩展的解决方案。","World Journal of Multidisciplinary Studies","2026-10-16T00:00:00Z",67,{"impact":17,"substance":90,"depth":91,"authority":167,"freshness":67,"relevant":21,"comment":210},"方法扎实、验证充分，但属单点应用研究，公共影响有限，适合作为技术前沿条目而非头条精选。",[212],{"name":206,"url":203},[97,150,214,215,216],"农业物联网","温室环境","小气候预测",{"code":38,"slug":39,"name":40,"path":41,"parent_code":35},[219,224],{"code":220,"slug":221,"name":222,"path":223,"parent_code":47},"I5","agricultural-iot","物联网传感与边缘计算","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-iot",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},{"status":49,"locked":14,"version":50,"method":51,"evidence":226},{"candidates":227,"agricultural_evidence":55},[228,234,240,247,253,257,264],{"code":220,"score":161,"supported":55,"title":229,"abstract":231,"tags":233},[230],"IoT",[232,230],"物联网",[214],{"code":38,"score":235,"supported":55,"title":236,"abstract":238,"tags":239},14,[237],"greenhouse",[69,237],[],{"code":31,"score":167,"supported":55,"title":241,"abstract":242,"tags":246},[],[63,243,244,163,245],"辣椒","黄瓜","cucumber",[],{"code":117,"score":173,"supported":55,"title":248,"abstract":249,"tags":252},[],[250,251],"精准农业","precision agriculture",[97],{"code":43,"score":173,"supported":55,"title":254,"abstract":255,"tags":256},[],[151,170],[150],{"code":258,"score":79,"supported":55,"title":259,"abstract":260,"tags":263},"H2",[],[261,262],"土壤","soil",[],{"code":265,"score":79,"supported":55,"title":266,"abstract":267,"tags":270},"H3",[],[268,269],"灌溉","irrigation",[],[272,273],"温室 小气候 LSTM 预测","尼日利亚 温室 IoT 黄瓜 番茄","温室小气候LSTM预测-5313",{"doi":8,"openalex_id":276,"authors":277,"venue":206,"cited_by_count":125,"oa_url":8,"card":288,"direction":132,"ingested_from":198},"W7222909837",[278,280,282,284,286],{"name":279,"orcid":8},"Siman Emmanuel*",{"name":281,"orcid":8},"Edward Onyebueke Agu",{"name":283,"orcid":8},"Musa Gani",{"name":285,"orcid":8},"Oladunjoye John Abiodun",{"name":287,"orcid":8},"Anagu Emmanuel John",{"tldr":289,"method":290,"finding":291,"direction":132,"opportunity":292},"提出物联网集成LSTM与Bi-LSTM加权集成框架，实时预测温室微气候。","IoT传感器采集温光湿数据，经预处理后用LSTM-Bi-LSTM加权集成建模。","Bi-LSTM优于LSTM，R²达0.9981，统计检验证实性能提升显著稳健。","可探索边缘端轻量化部署与多作物跨温室迁移学习，提升实时预测泛化性。","2026-10-19T23:30:13.295236Z",{"id":295,"title":296,"url":297,"summary":298,"summary_zh":299,"content":8,"source_name":300,"source_url":297,"published_at":301,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":302,"score_detail":303,"sources":306,"tags":308,"primary_topic":311,"secondary_topics":312,"classification":315,"search_phrases":344,"slug":347,"view_count":20,"doi":348,"paper":349,"created_at":369},4569,"SMART GREENHOUSE MEASUREMENT AND MONITORING SYSTEM BASED ON IOT AND COMPUTER VISION","https:\u002F\u002Fdoi.org\u002F10.46242\u002Fjstiuh.v83i05.6189","This paper presents a smart greenhouse measurement and monitoring system integrating IoT and computer vision for non-invasive plant growth observation and remote supervision. The proposed architecture combines environmental sensors, STM32-based data acquisition and control modules, a Raspberry Pi 4 with a USB camera, a lightweight deep learning-based stem segmentation module, Firebase Realtime Database, and a web application for monitoring and control. In the computer vision module, tomato plant images are processed directly on the Raspberry Pi to segment stem regions and estimate plant height from calibrated image measurements, enabling edge-based phenotypic monitoring without destructive sampling or manual measurement. Experimental results on a tomato cultivation prototype show that the system can collect temperature, air humidity, soil moisture, and light intensity data in real time while providing automated plant height estimation from RGB images. During the reported period, temperature ranged from 27.8°C to 38.8°C, air humidity from 40.2% to 77.0%, soil moisture from 52.0% to 82.4%, and plant height increased from 65.8 cm to 77.0 cm. The segmentation results demonstrate the feasibility of extracting stem-level structural information under practical greenhouse conditions, including complex background, leaf occlusion, and non-uniform illumination. The prototype confirms the potential of a low-cost smart greenhouse monitoring system and provides a practical basis for future expansion toward richer agronomic indicators, such as stem width, leaf area, canopy coverage, growth rate estimation, and automated environmental control.","本文提出了一种集成物联网与计算机视觉的智能温室测量与监测系统，用于非侵入式植物生长观测与远程监管。所提出的架构结合了环境传感器、基于STM32的数据采集与控制模块、配备USB摄像头的Raspberry Pi 4、基于轻量级深度学习的茎秆分割模块、Firebase实时数据库以及用于监测与控制的Web应用程序。在计算机视觉模块中，番茄植株图像直接在Raspberry Pi上处理，以分割茎秆区域并根据标定后的图像测量值估算株高，从而实现基于边缘的表型监测，无需破坏性采样或人工测量。在番茄栽培原型上的实验结果表明，该系统能够实时采集温度、空气湿度、土壤湿度和光照强度数据，同时基于RGB图像提供自动株高估算。在报告期内，温度范围为27.8°C至38.8°C，空气湿度为40.2%至77.0%，土壤湿度为52.0%至82.4%，株高从65.8 cm增长至77.0 cm。分割结果证明了在实际温室条件下（包括复杂背景、叶片遮挡和非均匀光照）提取茎秆级结构信息的可行性。该原型验证了低成本智能温室监测系统的潜力，并为未来扩展至更丰富的农艺指标（如茎宽、叶面积、冠层覆盖度、生长速率估算和自动化环境控制）提供了实践基础。","Journal of Science and Technology - IUH","2026-10-08T00:00:00Z",66,{"impact":17,"substance":304,"depth":19,"authority":13,"freshness":67,"relevant":21,"comment":305},20,"低成本IoT与边缘视觉融合的温室监测原型，方法可行、数据翔实，但属细分技术验证，产业影响有限。",[307],{"name":300,"url":297},[97,150,309,214,310],"计算机视觉","温室种植",{"code":38,"slug":39,"name":40,"path":41,"parent_code":35},[313,314],{"code":43,"slug":44,"name":45,"path":46,"parent_code":47},{"code":220,"slug":221,"name":222,"path":223,"parent_code":47},{"status":49,"locked":14,"version":50,"method":51,"evidence":316},{"candidates":317,"agricultural_evidence":55},[318,324,328,332,336,340],{"code":43,"score":319,"supported":55,"title":320,"abstract":322,"tags":323},19,[321],"computer vision",[151,309,170,321],[150,309],{"code":38,"score":161,"supported":55,"title":325,"abstract":326,"tags":327},[237],[69,237],[310],{"code":220,"score":161,"supported":55,"title":329,"abstract":330,"tags":331},[230],[232,230],[214],{"code":31,"score":79,"supported":55,"title":333,"abstract":334,"tags":335},[],[63,163],[],{"code":258,"score":79,"supported":55,"title":337,"abstract":338,"tags":339},[],[261,262],[],{"code":117,"score":21,"supported":14,"title":341,"abstract":342,"tags":343},[],[],[97],[345,346],"Raspberry Pi 温室监测","番茄 株高 图像识别","RaspberryPi温室监测-4569","10.46242\u002Fjstiuh.v83i05.6189",{"doi":348,"openalex_id":350,"authors":351,"venue":300,"cited_by_count":125,"oa_url":363,"card":364,"direction":132,"ingested_from":198},"W7221046739",[352,354,356,358,360],{"name":353,"orcid":8},"HOANG THI TU UYEN",{"name":355,"orcid":8},"NGUYEN HOANH",{"name":357,"orcid":8},"NGO THANH QUYEN",{"name":359,"orcid":8},"Lam Binh Minh",{"name":361,"orcid":362},"Nguyễn Trung Nhân","https:\u002F\u002Forcid.org\u002F0000-0001-9328-7693","https:\u002F\u002Fjst.iuh.edu.vn\u002Findex.php\u002Fjst-iuh\u002Farticle\u002Fdownload\u002F6189\u002F1131",{"tldr":365,"method":366,"finding":367,"direction":132,"opportunity":368},"集成物联网与计算机视觉的智能温室系统，实现非破坏性植物生长监测与远程管控。","环境传感器、STM32、树莓派4、USB摄像头、轻量茎分割模型、Firebase","原型可实时采集温湿度、土壤水分和光照，并从RGB图像估计株高，验证了边缘表型监测可行性。","可扩展至茎宽、叶面积、冠层覆盖度等更丰富农艺指标，并实现自动化环境控制闭环。","2026-10-10T23:30:11.420750Z",{"id":371,"title":372,"url":373,"summary":374,"summary_zh":375,"content":8,"source_name":376,"source_url":373,"published_at":377,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":302,"score_detail":378,"sources":380,"tags":382,"primary_topic":384,"secondary_topics":385,"classification":387,"search_phrases":402,"slug":405,"view_count":21,"doi":406,"paper":407,"created_at":430},4494,"Tomato ripeness classification with weight-based and texture-adaptive pooling in CNNs","https:\u002F\u002Fdoi.org\u002F10.64960\u002Feasr.2026.266223","This research introduces a weight-based mixed pooling and adaptive fuzzy pooling mechanism based on feature map texture analysis for attention mechanism Squeeze-and-Excitation Networks (SE) and Convolutional Neural Networks (CNNs) in tomato ripeness classification. Traditional CNNs and SE networks rely on deterministic pooling strategies that can be suboptimal for capturing diverse visual features such as texture and color gradients in fruit. A novel fuzzy pooling system is proposed that dynamically adjusts the fusion ratio of min and max pooling based on real-time texture analysis of feature maps via a Laplacian filter. This mechanism was integrated into Darknet53, ResNet34, and ResNet101 architectures, both directly and within the SE attention block. The method was evaluated on a five-class tomato ripeness dataset using 10-fold cross-validation against traditional and fixed-weight pooling methods. The fuzzy and mixed pooling approach demonstrated superior performance on certain models which include ESEnet: Darknet53 + SE (Fuzzy pool) achieved an average F1 score across folds of 0.8429, outperforming Darknet53 + baseline SE (Average pool) at 0.8350; Extended Net: ResNet34 + Fuzzy pool achieved an average F1 score across folds of 0.8305, outperforming baseline ResNet34 at 0.8130 and ResNet101 + Fuzzy pool achieved an average F1 score across folds of 0.8284, outperforming baseline ResNet101 at 0.8267. This improvement was realized with minimal increase in computational overhead. Specifically, measurements of model parameters (M) and FLOPs (GMACs) showed only a slight increase across all architectures after integrating SEnet into the system, and zero increase after integrating the Laplacian filter and fuzzy logic. The study concludes that logic-based adaptive pooling is a robust strategy that improves feature representation in CNNs, achieving a balance between classification accuracy and computational cost. These findings offer practical value for developing more reliable automated harvesting systems and provide a generalizable enhancement for a wide range of computer vision classification tasks.","本研究提出了一种基于特征图纹理分析的权重混合池化与自适应模糊池化机制，用于番茄成熟度分类中的注意力机制挤压激励网络（Squeeze-and-Excitation Networks, SE）和卷积神经网络（Convolutional Neural Networks, CNNs）。传统的CNN和SE网络依赖确定性池化策略，在捕捉果实纹理和颜色梯度等多样化视觉特征时可能并非最优。本文提出了一种新颖的模糊池化系统，通过拉普拉斯滤波器对特征图进行实时纹理分析，动态调整最小池化和最大池化的融合比例。该机制被集成到Darknet53、ResNet34和ResNet101架构中，既可直接集成，也可嵌入SE注意力模块内。该方法在五类番茄成熟度数据集上采用10折交叉验证，与传统池化及固定权重池化方法进行了对比评估。模糊池化和混合池化方法在部分模型上表现出优越性能，包括：ESEnet：Darknet53 + SE（模糊池化）在各折上的平均F1得分为0.8429，优于Darknet53 + 基线SE（平均池化）的0.8350；扩展网络：ResNet34 + 模糊池化在各折上的平均F1得分为0.8305，优于基线ResNet34的0.8130；ResNet101 + 模糊池化在各折上的平均F1得分为0.8284，优于基线ResNet101的0.8267。这一改进仅带来了极小的计算开销增加。具体而言，模型参数量（M）和浮点运算次数（GMACs）的测量结果表明，将SEnet集成到系统后所有架构仅有轻微增加，而集成拉普拉斯滤波器和模糊逻辑后则无任何增加。研究结论认为，基于逻辑的自适应池化是一种鲁棒策略，能够改善CNN中的特征表示，在分类精度和计算成本之间实现平衡。这些发现为开发更可靠的自动化收获系统提供了实用价值，并为广泛的计算机视觉分类任务提供了一种可推广的增强方法。","Engineering and Applied Science Research","2026-10-07T00:00:00Z",{"impact":67,"substance":304,"depth":91,"authority":17,"freshness":167,"relevant":21,"comment":379},"提出基于纹理自适应模糊池化的CNN改进方法，在番茄成熟度分类上取得小幅但稳定的F1提升且算力开销几乎不增，对自动采摘与农业视觉识别有参考价值，但属细分方法类论文，影响面有限。",[381],{"name":376,"url":373},[97,150,63,153,383],"自动采摘",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[386],{"code":43,"slug":44,"name":45,"path":46,"parent_code":47},{"status":49,"locked":14,"version":50,"method":51,"evidence":388},{"candidates":389,"agricultural_evidence":55},[390,394,398],{"code":31,"score":161,"supported":55,"title":391,"abstract":392,"tags":393},[163],[63,163],[63],{"code":43,"score":67,"supported":55,"title":395,"abstract":396,"tags":397},[],[309,321],[150,153],{"code":117,"score":21,"supported":14,"title":399,"abstract":400,"tags":401},[],[],[97],[403,404],"番茄 成熟度 分类 卷积神经网络","Darknet53 ResNet 番茄 成熟度","番茄成熟度分类卷积神经网络-4494","10.64960\u002Feasr.2026.266223",{"doi":406,"openalex_id":408,"authors":409,"venue":376,"cited_by_count":125,"oa_url":424,"card":425,"direction":132,"ingested_from":198},"W7220840158",[410,412,414,416,419,421],{"name":411,"orcid":8},"Rut Yatigul",{"name":413,"orcid":8},"Wisanu Jitviriya",{"name":415,"orcid":8},"Eiji Hayashi",{"name":417,"orcid":418},"Aran Blattler","https:\u002F\u002Forcid.org\u002F0000-0002-8684-6239",{"name":420,"orcid":8},"Watcharin Tangsuksant",{"name":422,"orcid":423},"Teppakorn Sittiwanchai","https:\u002F\u002Forcid.org\u002F0000-0001-9653-8750","https:\u002F\u002Fph01.tci-thaijo.org\u002Findex.php\u002Feasr\u002Farticle\u002Fdownload\u002F266223\u002F176569",{"tldr":426,"method":427,"finding":428,"direction":196,"opportunity":429},"提出基于纹理自适应模糊池化机制，提升CNN番茄成熟度分类性能。","用Laplacian滤波分析特征图纹理，动态调整最大最小池化融合比，集成到Dar","模糊池化在多个模型上F1提升，且计算开销几乎不增加。","可探索自适应池化在其它果蔬成熟度或田间复杂光照下的泛化能力。","2026-10-09T23:30:09.993787Z",{"id":432,"title":433,"url":434,"summary":435,"summary_zh":8,"content":436,"source_name":437,"source_url":8,"published_at":438,"category":439,"cover_url":8,"hotness":13,"is_selected":14,"score":440,"score_detail":441,"sources":444,"tags":446,"primary_topic":450,"secondary_topics":451,"classification":453,"search_phrases":473,"slug":476,"view_count":465,"doi":8,"paper":8,"created_at":477},4459,"中国农科院环发所构建知识与数据双驱动模型精准预测温室番茄生长态势","https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572492.shtm","中国农业科学院农业环境与可持续发展研究所设施植物环境工程创新团队以番茄为对象，构建知识与数据双驱动的尺度自适应时序卷积混合模型，根据番茄不同生育期动态调整时序特征提取尺度，结合机理模型合成数据预训练与生理阈值约束，实现番茄不同器官干物质及叶面积指数的精准预测；研究成果发表在Computers and Electronics in Agriculture；仅依靠常规温室环境数据即可无损估算生长态势，对推动温室智能环控、产量预测具重要应用价值。","[![Image 1: 科学网新闻频道](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Fnews.jpg)](https:\u002F\u002Fnews.sciencenet.cn\u002F)\n\n[生命科学](https:\u002F\u002Fwww.sciencenet.cn\u002Flife\u002F) | [医学科学](https:\u002F\u002Fwww.sciencenet.cn\u002Fmedicine\u002F) | [化学科学](https:\u002F\u002Fwww.sciencenet.cn\u002Fchemistry\u002F) | [工程材料](https:\u002F\u002Fwww.sciencenet.cn\u002Fmaterial\u002F) | [信息科学](https:\u002F\u002Fwww.sciencenet.cn\u002Finformation\u002F) | [地球科学](https:\u002F\u002Fwww.sciencenet.cn\u002Fearth\u002F) | [数理科学](https:\u002F\u002Fwww.sciencenet.cn\u002Fmathematics\u002F) | [管理综合](https:\u002F\u002Fwww.sciencenet.cn\u002Fpolicy\u002F)[站内规定](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-45-1064777.html) | [手机版](https:\u002F\u002Fwap.sciencenet.cn\u002F)\n\n[首页](https:\u002F\u002Fwww.sciencenet.cn\u002F) | [新闻](https:\u002F\u002Fnews.sciencenet.cn\u002F) | [博客](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog.php) | [院士](https:\u002F\u002Fnews.sciencenet.cn\u002Fys\u002F) | [人才](https:\u002F\u002Ftalent.sciencenet.cn\u002F) | [会议](https:\u002F\u002Fmeeting.sciencenet.cn\u002F) | [基金·项目](https:\u002F\u002Ffund.sciencenet.cn\u002F) | [论文](https:\u002F\u002Fpaper.sciencenet.cn\u002F) | [绘图](https:\u002F\u002Fvisual-web.sciencenet.cn\u002F) | [视频·直播](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog.php?mod=video) | [小柯机器人](https:\u002F\u002Fpaper.sciencenet.cn\u002FAInews) | [医学科普](https:\u002F\u002Fyxkp.sciencenet.cn\u002F)\n\n作者：李晨,王佳 来源：中国科学报 发布时间：2026\u002F10\u002F4 22:09:23 选择字号：小 中 大 双驱动模型可精准预测温室番茄生长态势 近日，中国农业科学院农业环境与可持续发展研究所设施植物环境工程创新团队构建了知识与数据双驱动模型，实现温室番茄生长的高精度无损预测，相关研究成果发表在《农业计算机与电子学》（_**Computers and Electronics in Agriculture**_）上。 作物生长精准估算是温室智能控制的核心前提。然而，传统机理模型参数获取难、泛化有限，而纯数据驱动模型常因缺乏生理约束致预测失真。因此，如何融合环境时序数据与作物生长规律，实现高精度且生理合理的生长预测仍是关键挑战。 科研团队以番茄为对象，提出一种知识与数据双驱动的尺度自适应时序卷积混合模型。该模型根据番茄不同生育期动态调整时序特征提取尺度，结合机理模型合成数据预训练与生理阈值约束，实现番茄不同器官干物质及叶面积指数的精准预测。该模型依靠常规温室环境数据即可无损估算生长态势，对推动温室智能环控、产量预测具有重要应用价值。 相关论文信息：https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.compag.2026.112281 版权声明：凡本网注明“来源：中国科学报、科学网、科学新闻杂志”的所有作品，网站转载，请在正文上方注明来源和作者，且不得对内容作实质性改动；微信公众号、头条号等新媒体平台，转载请联系授权。邮箱：shouquan@stimes.cn。\n\n打印 发E-mail给：\n\n以下评论只代表网友个人观点，不代表科学网观点。\n\n 无标题页 \n\n[![Image 2](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Fnewcomm.gif)](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtml\u002Fcomment.aspx?id=572492)\n\n相关新闻 相关论文\n*   1\n*   [《科学》（20261001出版）一周论文导读](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572487.shtm)\n\n*   2\n*   [2026年度湖南省科学技术奖提名工作启动](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572486.shtm)\n\n*   3\n*   [河南中央引导地方科技发展资金项目结项验收结果公示](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572485.shtm)\n\n*   4\n*   [江西组织开展2027年度自然科学基金项目申报工作](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572484.shtm)\n\n*   5\n*   [中国人民大学收到一校友捐资5.03亿](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572483.shtm)\n\n*   6\n*   [2026年度辽宁省科学技术奖提名工作启动](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572482.shtm)\n\n*   7\n*   [北京市自然科学基金资助项目验收结论公布](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572481.shtm)\n\n*   8\n*   [湖北省科技计划项目2026年第六批验收结果公示](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572480.shtm)\n\n*   1\n*   [草酸二甲酯加氢催化剂研究：解锁三类高价值化学品定向合成](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F202610318254299156239.shtm)\n\n*   2\n*   [杨孟昊研究员：人工智能赋能能源材料研究——从经典机器学习到大模型的技术演进与应用](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F2026103174511786156238.shtm)\n\n*   3\n*   [西北工业大学卢乾波教授：系统梳理压电驱动器种类，为其应用提供“知识地图”](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F202610317332521156234.shtm)\n\n*   4\n*   [异辛醇氧化工艺优化：混合建模策略实现节能降碳双提升](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F202610316823146156233.shtm)\n\n*   5\n*   [从物理智能到分子设计：跨尺度机制学习推动药物发现](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F202610316233349156232.shtm)\n\n*   6\n*   [流场+磁场双驱动—等离子体固氮技术](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F2026103151851943156231.shtm)\n\n*   7\n*   [根际细菌的“生化部队”能让植物更耐旱吗？](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F2026103105344865156230.shtm)\n\n*   8\n*   [从4000小时到更长寿命：解密PEM电解槽多孔传输层的设计哲学与产业博弈](https:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlpaper\u002F2026\u002F10\u002F2026101231244224156228.shtm)\n\n[图片新闻](https:\u002F\u002Fnews.sciencenet.cn\u002Fimgnews.aspx)\n\n[![Image 3: 他们为二维滑移铁电体装原子层“方向盘”](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fupload\u002Fnews\u002Fimages\u002F2026\u002F10\u002F20261061524217710.png)](https:\u002F\u002Fnews.sciencenet.cn\u002Fhttps:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572520.shtm)[![Image 4: 这种紫花苜蓿新种质可同时耐受两种除草剂](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fupload\u002Fnews\u002Fimages\u002F2026\u002F10\u002F2026104223520990.jpg)](https:\u002F\u002Fnews.sciencenet.cn\u002Fhttps:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F10\u002F572493.shtm)\n[![Image 5: “鹏城首发星”成功入轨](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fupload\u002Fnews\u002Fimages\u002F2026\u002F9\u002F20269231515219900.png)](https:\u002F\u002Fnews.sciencenet.cn\u002Fhttps:\u002F\u002Fnews.sciencenet.cn\u002Fhtmlnews\u002F2026\u002F9\u002F571953.shtm)[![Image 6: 研究揭示甲烷部分氧化制合成气驱动因素](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fupload\u002Fnews\u002Fimages\u002F2026\u002F9\u002F2026917125225680.jpg)](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F9\u002F571610.shtm)\n[>>更多](https:\u002F\u002Fnews.sciencenet.cn\u002Fimgnews.aspx)\n\n[一周新闻排行](https:\u002F\u002Fnews.sciencenet.cn\u002Fzphnews.aspx)\n*   1\n*   [2026年诺贝尔生理学或医学奖揭晓！](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572511.shtm)\n\n*   2\n*   [刚刚，2026年诺贝尔物理奖揭晓！](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572524.shtm)\n\n*   3\n*   [刚刚，2026年诺贝尔化学奖揭晓！](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572558.shtm)\n\n*   4\n*   [中国学者创立新体系被增设进入IFAC十二大技术领域](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572427.shtm)\n\n*   5\n*   [直播回放｜2026诺贝尔物理学奖解读](https:\u002F\u002Fnews.sciencenet.cn\u002Fhttps:\u002F\u002Fweibo.com\u002Fl\u002Fwblive\u002Fp\u002Fshow\u002F1022:2321325346726952435780)\n\n*   6\n*   [诺贝尔物理学奖揭晓在即：凝聚态物理研究受关注](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572459.shtm)\n\n*   7\n*   [施郁：今年诺贝尔物理学奖，我为何看好他们](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572473.shtm)\n\n*   8\n*   [美国白蛾大暴发，北京释放11亿头“天敌”！](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572445.shtm)\n\n*   9\n*   [诺贝尔生理学或医学奖今年花落谁家？](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572448.shtm)\n\n*   10\n*   [华人物理学家叶军获“诺奖风向标”沃尔夫物理学奖](https:\u002F\u002Fnews.sciencenet.cn\u002F\u002Fhtmlnews\u002F2026\u002F10\u002F572453.shtm)\n\n[编辑部推荐博文](http:\u002F\u002Fblog.sciencenet.cn\u002Fblog.php?mod=recommend)\n*   ![Image 7](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [封面文章|《水生生物学报》：微藻蛋白的研究进展](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-45-1554679.html)\n\n*   ![Image 8](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [科学网2026年8月十佳博文榜单公布！](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-45-1551789.html)\n\n*   ![Image 9](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [化学反应中的“神奇效应”](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-3380716-1555615.html)\n\n*   ![Image 10](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [主流同行评审模式盘点，投稿前没了解清楚？难怪初审总出状况](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-3574014-1555614.html)\n\n*   ![Image 11](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [柔性电子重塑未来科技形态](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-3496796-1555605.html)\n\n*   ![Image 12](https:\u002F\u002Fnews.sciencenet.cn\u002Fimages\u002Ft11.gif)\n*   [即将截止 | 填写问卷，告诉我们您对选刊工具的使用偏好，更有机会获得 ...](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog-3411312-1555592.html)\n\n*   [更多>>](https:\u002F\u002Fblog.sciencenet.cn\u002Fblog.php?mod=recommend)\n\n[关于我们](http:\u002F\u002Fwww.sciencenet.cn\u002Faboutus.aspx) | [网站声明](http:\u002F\u002Fwww.sciencenet.cn\u002Fshengming.aspx) | [服务条款](http:\u002F\u002Fwww.sciencenet.cn\u002Ftiaokuang.aspx) | [联系方式](http:\u002F\u002Fwww.sciencenet.cn\u002Fcontact.aspx) | [举报](https:\u002F\u002Fwww.sciencenet.cn\u002Fjubao.aspx) | [中国科学报社](https:\u002F\u002Fstimes.sciencenet.cn\u002F)  \n[京ICP备07017567号-12](http:\u002F\u002Fwww.beian.miit.gov.cn\u002F)互联网新闻信息服务许可证10120230008 京公网安备 11010802032783  \nCopyright @ 2007-2026 中国科学报社 All Rights Reserved  \n地址：北京市海淀区中关村南一条乙三号 电话：010-62580783","科学网 中国农业科学院 2026-10-04","2026-10-04T00:00:00Z","报道",80,{"impact":442,"substance":90,"depth":91,"authority":235,"freshness":79,"relevant":21,"comment":443},22,"国家级科研机构在智慧农业领域的模型方法突破，方法新颖、结论可靠，对温室智能环控与产量预测有应用价值，值得进入每日精选。",[445],{"name":437,"url":434},[97,150,447,448,449],"作物生长模型","设施番茄","温室环控",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[452],{"code":38,"slug":39,"name":40,"path":41,"parent_code":35},{"status":49,"locked":14,"version":50,"method":51,"evidence":454},{"candidates":455,"agricultural_evidence":55},[456,460,464,469],{"code":31,"score":17,"supported":55,"title":457,"abstract":458,"tags":459},[63],[63],[448],{"code":38,"score":61,"supported":55,"title":461,"abstract":462,"tags":463},[69],[69],[],{"code":43,"score":465,"supported":14,"title":466,"abstract":467,"tags":468},2,[],[],[447,150],{"code":117,"score":21,"supported":14,"title":470,"abstract":471,"tags":472},[],[],[97],[474,475],"中国农科院环发所 温室番茄 生长预测","知识与数据双驱动 番茄 干物质","中国农科院环发所温室番茄生长预测-4459","2026-10-09T00:04:56.060677Z"]