Query:
学者姓名:吴传宇
Refining:
Year
Type
Indexed by
Source
Complex
Co-Author
Language
Clean All
Abstract :
本实用新型提供了一种爬树修枝机器人,包括爬树机构,所述爬树机构包括四个驱动电机、四个行走轮、两根电机固定杆、两根伸缩杆和四个铰链,每个所述驱动电机带动一个所述行走轮转动,所述行走轮倾斜于水平面设置,每两个所述驱动电机固定在一根所述电机固定杆的内侧,两根所述电机固定杆平行设置,两根所述电机固定杆与两根所述伸缩杆组成矩形,相邻的所述电机固定杆与所述伸缩杆之间通过所述铰链连接。能有效防止修枝过程中出现侧翻的情况。
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 许汉良 , 洪启睿 , 杜远飞 et al. 爬树修枝机器人 : CN202421394438.8[P]. | 2024-06-18 . |
| MLA | 许汉良 et al. "爬树修枝机器人" : CN202421394438.8. | 2024-06-18 . |
| APA | 许汉良 , 洪启睿 , 杜远飞 , 熊盛佳 , 吴章云 , 叶智国 et al. 爬树修枝机器人 : CN202421394438.8. | 2024-06-18 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
针对传统方法在复杂地形和相似作物背景下难以精确识别茶园区域的问题,文章提出了一种基于增强编码器与多级特征融合的方法,旨在开发一种名为Tea-UNet的新型语义分割算法,专门用于从无人机获取的高分辨率影像中高效、准确地提取茶园区域。通过增强编码器,采用RepVit架构,结合深度可分离卷积和结构重参数化技术,在保持计算效率的同时增强了全局特征建模能力。此外,引入了基于Transformer的多级多尺度特征融合模块(MFT),通过多头交叉注意力机制实现复杂的特征融合,有效缩小了浅层与深层特征间的语义差距。解码器部分则设计了多级注意力引导的特征融合模块(MAF),通过连续的注意力机制更新,精炼特征表示,减少不相关信息干扰。实验结果表明,Tea-UNet模型在茶园区域提取任务上取得了显著的效果,不仅提高了分割精度,还增强了对细长或形状不规则目标的识别能力,该研究为茶园管理和可持续发展提供了强有力的技术支持。
Keyword :
增强编码器 增强编码器 多级特征融合 多级特征融合 无人机影像 无人机影像 茶园提取 茶园提取 语义分割 语义分割
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 熊盛佳 , 叶智国 , 吴章云 et al. Tea-UNet茶园无人机影像语义分割算法研究 [J]. | 信息技术与信息化 , 2025 , 6 (05) : 17-22 . |
| MLA | 熊盛佳 et al. "Tea-UNet茶园无人机影像语义分割算法研究" . | 信息技术与信息化 6 . 05 (2025) : 17-22 . |
| APA | 熊盛佳 , 叶智国 , 吴章云 , 吴传宇 . Tea-UNet茶园无人机影像语义分割算法研究 . | 信息技术与信息化 , 2025 , 6 (05) , 17-22 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
随着计算机视觉技术在自动驾驶、智能监控等领域的广泛应用,雨天环境因自身特性显著增加了图像去雨的难度,导致图像质量下降,这一问题亟待解决。为应对现有方法在提取关键雨条特征和避免信息冗余上的不足,文章提出了一种基于多尺度特征学习和注意力机制的密集连接网络(CMADNet),用于单幅图像的去雨处理。CMADNet 结合多尺度分割注意力(MSA)模块和多尺度特征学习(MFL)模块,在密集连接框架中可以有效提取雨条特征,同时抑制冗余信息。实验结果表明,CMADNet在多个基准数据集上均实现了先进的性能,显著提高了PSNR和SSIM指标。与现有方法相比,该方法具备更高的计算效率和泛化能力。验证了其在恶劣天气下计算机视觉任务中的应用潜力。
Keyword :
图像去雨 图像去雨 多尺度分割注意力 多尺度分割注意力 多尺度特征学习 多尺度特征学习 密集连接网络 密集连接网络
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 吴章云 , 吴传宇 , 叶智国 et al. 一种基于密集连接网络的图像去雨方法 [J]. | 信息技术与信息化 , 2025 , 5 (05) : 130-134 . |
| MLA | 吴章云 et al. "一种基于密集连接网络的图像去雨方法" . | 信息技术与信息化 5 . 05 (2025) : 130-134 . |
| APA | 吴章云 , 吴传宇 , 叶智国 , 熊盛佳 , 王硕 , 林凯旋 . 一种基于密集连接网络的图像去雨方法 . | 信息技术与信息化 , 2025 , 5 (05) , 130-134 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
室内环境中,因摄像机拍摄角度多变、光线条件不稳定,导致老年人跌倒状况难以精准识别,为此,文章提出了一种基于YOLOv8s改进的RRS-YOLO算法。在Backbone层,采用基于GELAN的模块改进C2f模块,以获取更丰富特征信息,同时减少计算量和内存消耗。在Neck层,采用重参数加权双向特征金字塔网络,融合不同特征层,提升检测准确性和鲁棒性。检测头部分引入DWConv和SaE自编码器,关注小目标和遮挡目标,提高整体性能。实验显示,RRS-YOLO算法参数量降低19.4%,计算量降低9.2%,mAP@0.5提升2.1%。该算法不仅降低了模型参数量,还显著提升了跌倒检测速度与精度,有助于完善养老服务体系,提升老年人生活安全性和幸福感。
Keyword :
RGELAN RGELAN RSBiPAN RSBiPAN SD SD YOLOv8s YOLOv8s 跌倒检测 跌倒检测
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 叶智国 , 熊盛佳 , 吴章云 et al. 基于YOLOv8s改进的居家跌倒行为检测算法 [J]. | 信息技术与信息化 , 2025 , 5 (05) : 3-7 . |
| MLA | 叶智国 et al. "基于YOLOv8s改进的居家跌倒行为检测算法" . | 信息技术与信息化 5 . 05 (2025) : 3-7 . |
| APA | 叶智国 , 熊盛佳 , 吴章云 , 杜远飞 , 吴传宇 , 陈仕国 . 基于YOLOv8s改进的居家跌倒行为检测算法 . | 信息技术与信息化 , 2025 , 5 (05) , 3-7 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
Path planning for unmanned aerial vehicles (UAVs) in mountainous environments requires satisfying terrain clearance and obstacle avoidance constraints while optimizing path length, flight time, and energy consumption. To address these challenges, this paper proposes EC-MOPSO (Epsilon-dominance and Crowding-distance-based Multi-Objective Particle Swarm Optimization). Inspired by the principle of symmetry, the algorithm integrates an adaptive parameter adjustment mechanism with a epsilon- dominance-crowding archiving strategy to balance global exploration and local exploitation through spatially symmetric archive management. A safety-repairable B-spline trajectory model ensures smooth and feasible flight paths under complex terrain conditions. Simulation results show that EC-MOPSO reduces path length by 10-40%, improves normalized hypervolume by over 25%, and decreases performance variance by 20-25%, confirming faster convergence and higher robustness compared with representative multi-objective optimization approaches. Ablation studies further verify that both the adaptive parameter mechanism and the epsilon- dominance-crowding strategy significantly enhance convergence stability and overall optimization performance. Overall, EC-MOPSO provides an adaptive and reliable optimization framework for generating efficient, safe, and energy-aware UAV trajectories in real-world mountainous rescue missions.
Keyword :
emergency rescue emergency rescue epsilon-dominance epsilon-dominance mountainous environments mountainous environments multi-objective particle swarm optimization multi-objective particle swarm optimization repairable operator repairable operator unmanned aerial vehicles unmanned aerial vehicles
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | Zou, Wenxing , Xu, Hang , Chen, Chuze et al. An Enhanced MOPSO Algorithm for Multi-Objective UAV Path Planning in Mountainous Environments [J]. | SYMMETRY-BASEL , 2025 , 17 (11) . |
| MLA | Zou, Wenxing et al. "An Enhanced MOPSO Algorithm for Multi-Objective UAV Path Planning in Mountainous Environments" . | SYMMETRY-BASEL 17 . 11 (2025) . |
| APA | Zou, Wenxing , Xu, Hang , Chen, Chuze , Wu, Chuanyu . An Enhanced MOPSO Algorithm for Multi-Objective UAV Path Planning in Mountainous Environments . | SYMMETRY-BASEL , 2025 , 17 (11) . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
针对传统卷积神经网络难以有效提取人脸表情特征,易出现梯度异常而导致识别准确率不佳等问题,建立了一种基于残差学习策略的深度卷积神经网络,并引入稀疏自注意力机制和形变卷积核改进网络性能,改进后的网络命名为Cross_PainRec_Net。根据科学的疼痛评估法建立了三分类的痛苦表情数据库。利用上述数据库设置网络性能验证实验、消融实验和网络性能对比实验,结果表明Cross_PainRec_Net的痛苦表情识别准确率达到94.2%,比经典的图片分类网络Mobile Net、VggNet、ResNet分别高出46.5%、19%和3.2%,消融实验证明稀疏自注意力模块和形变卷积模块对网络均有增益作用。
Keyword :
形变卷积 形变卷积 残差学习 残差学习 深度卷积神经网络 深度卷积神经网络 痛苦表情特征 痛苦表情特征 稀疏自注意力机制 稀疏自注意力机制
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 胡卓成 , 陈仕国 , 叶智国 et al. 基于深度卷积神经网络的痛苦表情识别算法 [J]. | 计算机仿真 , 2025 , 42 (08) : 484-490 . |
| MLA | 胡卓成 et al. "基于深度卷积神经网络的痛苦表情识别算法" . | 计算机仿真 42 . 08 (2025) : 484-490 . |
| APA | 胡卓成 , 陈仕国 , 叶智国 , 吴传宇 . 基于深度卷积神经网络的痛苦表情识别算法 . | 计算机仿真 , 2025 , 42 (08) , 484-490 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
The aim of this study was to investigate the relationship model between the ultrasonic propagation velocity and the mechanical property of side-pressed laminated bamboo lumber (SPLBL), as well as to provides a theoretical basis for non-destructive ultrasound testing on the mechanical properties of SPLBL. Taking SPLBL as the research object and using a microsecond ultrasound measuring instrument, the ultrasonic propagation velocity in SPLBL and small specimens was obtained through the ultrasonic propagation measurement test, and a contour distribution map of the ultrasonic propagation speed of SPLBL was drawn. After that, the flexural elastic modulus and flexural strength of the small specimens were obtained through the three-point bending test, and regression analysis was performed on the relationship between ultrasonic propagation velocity and flexural modulus and flexural strength. There were significant differences in the contour distribution of ultrasonic wave velocity obtained for each SPLBL sample, indicating that there may be differences in the gluing effect, resulting in certain differences in the mechanical properties. The higher the velocity of ultrasonic wave propagation, the better the gluing effect or mechanical properties of SPLBL. A good correlation was found between ultrasonic wave velocity and the flexural elastic modulus and flexural strength of small specimens (coefficient of determination R2 = 0.52 and 0.46, respectively). This shows that ultrasonic wave velocity can reasonably predict and evaluate the flexural elastic modulus and flexural strength of small specimens. A good correlation was also found between the dynamic elastic modulus and flexural elastic modulus of small specimens, with an R2 for both of 0.55. This shows that it is also possible to reasonably predict and evaluate the flexural elastic modulus by measuring the dynamic elastic modulus of small specimens. Non-destructive ultrasonic testing technology is a potentially effective means to reasonably predict and evaluate the mechanical properties of SPLBL.
Keyword :
laminated bamboo lumber laminated bamboo lumber mechanical property mechanical property non-destructive testing non-destructive testing ultrasonic ultrasonic velocity velocity
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | Liu, Fenglu , Xiao, Jiawei , Zhao, Yang et al. Mechanical Properties of Side-Pressed Laminated Bamboo Lumber Based on Ultrasonic NDT Technology [J]. | FOREST SCIENCE AND TECHNOLOGY , 2025 . |
| MLA | Liu, Fenglu et al. "Mechanical Properties of Side-Pressed Laminated Bamboo Lumber Based on Ultrasonic NDT Technology" . | FOREST SCIENCE AND TECHNOLOGY (2025) . |
| APA | Liu, Fenglu , Xiao, Jiawei , Zhao, Yang , Wu, Chuanyu , Chen, Wenhao , Wang, Qinhui . Mechanical Properties of Side-Pressed Laminated Bamboo Lumber Based on Ultrasonic NDT Technology . | FOREST SCIENCE AND TECHNOLOGY , 2025 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
针对当前咖啡豆烘焙状态识别耗时耗力的问题,提出一种基于局部注意力增强的轻量化网络Rep—FdNet,实现对烘焙程度的实时监控。该网络采用一种全新的分频模块,增强网络对于局部特征的关注,有助于提高网络区分高频和低频信息的能力。在推理阶段采用重参数化方法将三分支结构融合成单路结构,在保证网络准确率的同时,加快网络推理速度并减少内存占用。试验结果表明:所提出的Rep—FdNet在分类准确率上达到98.2%,满足分类的需求;在计算量、参数量以及内存占用上分别仅有25.80 M、1.02 M和2.75 MB,有效解决计算资源有限的问题;在推理速度上达到124.99帧/s,满足工业应用上实时分类的要求。该轻量化网络Rep—FdNet能够识别咖啡豆烘焙程度,降低咖啡烘焙的人工成本和操作难度。
Keyword :
咖啡豆烘焙 咖啡豆烘焙 图像分频 图像分频 深度学习 深度学习 轻量化网络 轻量化网络 重参数化 重参数化
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 杜远飞 , 郭仕豪 , 吴传宇 et al. 基于轻量级网络设计的咖啡豆烘焙状态识别研究 [J]. | 中国农机化学报 , 2025 , 46 (09) : 212-219 . |
| MLA | 杜远飞 et al. "基于轻量级网络设计的咖啡豆烘焙状态识别研究" . | 中国农机化学报 46 . 09 (2025) : 212-219 . |
| APA | 杜远飞 , 郭仕豪 , 吴传宇 , 闫轩旭 . 基于轻量级网络设计的咖啡豆烘焙状态识别研究 . | 中国农机化学报 , 2025 , 46 (09) , 212-219 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
本实用新型提供了一种用于暗穴植树的松土挖穴刀,包括手柄和松土刀,所述手柄呈T型,所述手柄的底端连接所述松土刀的顶端,所述松土刀包括水平截面形状为X形的X型刀体,所述X型刀体中部镂空设置,使所述X型刀体包含四个竖直设置的刀条,且四个所述刀条的底端汇聚成至一个点,所述刀条的外侧面上设有刀刃。该松土挖穴刀能轻松省力的手动进行暗穴松土挖穴作业,便于暗穴植树时提高工作效率,节省人工,且便于携带。
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 熊盛佳 , 吴传宇 , 叶智国 et al. 一种用于暗穴植树的松土挖穴刀 : CN202420429978.9[P]. | 2024-03-06 . |
| MLA | 熊盛佳 et al. "一种用于暗穴植树的松土挖穴刀" : CN202420429978.9. | 2024-03-06 . |
| APA | 熊盛佳 , 吴传宇 , 叶智国 , 吴章云 , 章程杰 . 一种用于暗穴植树的松土挖穴刀 : CN202420429978.9. | 2024-03-06 . |
| Export to | NoteExpress RIS BibTex |
Version :
Abstract :
本发明提供了一种爬树修枝机器人,包括爬树机构,所述爬树机构包括四个驱动电机、四个行走轮、两根电机固定杆、两根伸缩杆和四个铰链,每个所述驱动电机带动一个所述行走轮转动,所述行走轮倾斜于水平面设置,每两个所述驱动电机固定在一根所述电机固定杆的内侧,两根所述电机固定杆平行设置,两根所述电机固定杆与两根所述伸缩杆组成矩形,相邻的所述电机固定杆与所述伸缩杆之间通过所述铰链连接。能有效防止修枝过程中出现侧翻的情况。
Cite:
Copy from the list or Export to your reference management。
| GB/T 7714 | 吴传宇 , 董楸煌 , 陈勇 et al. 爬树修枝机器人 : CN202410782085.7[P]. | 2024-06-18 . |
| MLA | 吴传宇 et al. "爬树修枝机器人" : CN202410782085.7. | 2024-06-18 . |
| APA | 吴传宇 , 董楸煌 , 陈勇 , 洪启睿 , 马祥庆 , 曹光球 et al. 爬树修枝机器人 : CN202410782085.7. | 2024-06-18 . |
| Export to | NoteExpress RIS BibTex |
Version :
Export
| Results: |
Selected to |
| Format: |