In positive-ion fast atom bombardment (FAB) mass spectrometry, when mono- and di- saccharides are mixed with an appropriate amount of NH4Cl, a highly abundan peak [M+NH4]+appers in FAB mass spectra . From the adduct ...In positive-ion fast atom bombardment (FAB) mass spectrometry, when mono- and di- saccharides are mixed with an appropriate amount of NH4Cl, a highly abundan peak [M+NH4]+appers in FAB mass spectra . From the adduct ion [M+NH4]+, the molecular weights of mono- and di- saccharides can be determined definitively展开更多
森林火点检测在林火应急救援中起着至关重要的作用.鉴于现有模型在样本质量、多尺度检测以及多视角图像泛化能力方面存在不足,以YOLOv7为基础,提出一种森林火点目标检测方法FFD-YOLO(forest fire detection based on YOLO).首先,构建多...森林火点检测在林火应急救援中起着至关重要的作用.鉴于现有模型在样本质量、多尺度检测以及多视角图像泛化能力方面存在不足,以YOLOv7为基础,提出一种森林火点目标检测方法FFD-YOLO(forest fire detection based on YOLO).首先,构建多视角可见光图像森林火灾高点检测数据集FFHPV(forest fire of high point view),旨在增强模型对多视角火点知识的学习能力;其次,引入全维动态卷积,构建空间金字塔池化层(OD-SPP),以此提升模型针对多视角数据的火点特征提取能力;最后,引入具有动态非单调聚焦机制的边界框定位损失函数Wise-IoU(wise intersection over union),降低低质量数据对模型精度的影响,提高小目标火点的检测能力.实验结果表明:所提出的FFD-YOLO方法相较于YOLOv7,精度提高3.9%,召回率提高3.7%,均值平均精度提高4.0%,F1分数提高0.038;同时,在与YOLOv5、YOLOv8、DDQ(dense distinct query)、DINO(detection transformer with improved denoising anchor boxes)、Faster R-CNN、Sparse R-CNN、Mask R-CNN、FCOS和YOLOX的对比实验中,FFD-YOLO具有最高的精度75.3%、召回率73.8%、均值平均精度77.6%和F1分数0.745,验证了该方法的可行性与有效性.展开更多
Ultrasound plays an important role not only in preoperative diagnosis but also in intraoperative guidance for liver surgery.Intraoperative ultrasound(IOUS)has become an indispensable tool for modern liver surgeons,esp...Ultrasound plays an important role not only in preoperative diagnosis but also in intraoperative guidance for liver surgery.Intraoperative ultrasound(IOUS)has become an indispensable tool for modern liver surgeons,especially for minimally invasive surgeries,partially substituting for the surgeon’s hands.In fundamental mode,Doppler mode,contrast enhancement,elastography,and real-time virtual sonography,IOUS can provide additional real-time information regarding the intrahepatic anatomy,tumor site and characteristics,macrovascular invasion,resection margin,transection plane,perfusion and outflow of the remnant liver,and local ablation efficacy for both open and minimally invasive liver resections.Identification and localization of intrahepatic lesions and surrounding structures are crucial for performing liver resection,preserving the adjacent vital vascular and bile ducts,and sparing the functional liver parenchyma.Intraoperative ultrasound can provide critical information for intraoperative decision-making and navigation.Therefore,all liver surgeons must master IOUS techniques,and IOUS should be included in the training of modern liver surgeons.Further investigation of the potential benefits and advances in these techniques will increase the use of IOUS in modern liver surgeries worldwide.This study comprehensively reviews the current use of IOUS in modern liver surgeries.展开更多
为满足多数工业场景下钢板表面缺陷检测的需求,针对钢板表面缺陷检测准确率低及小目标缺陷检测率低等问题,文中提出了一种基于改进YOLOv5(You Only Look Once version 5)的钢板表面缺陷检测算法。在YOLOv5的基础上将CBAM(Convolution Bl...为满足多数工业场景下钢板表面缺陷检测的需求,针对钢板表面缺陷检测准确率低及小目标缺陷检测率低等问题,文中提出了一种基于改进YOLOv5(You Only Look Once version 5)的钢板表面缺陷检测算法。在YOLOv5的基础上将CBAM(Convolution Block Attention Module)注意力模块嵌入到主干网络中,提高网络检测精度。加入上下文增强模块,提高了算法对小目标的检测性能。使用NWD(Normalized Wasserstein Distance)度量标准代替原YOLOv5中的IoU(Intersection over Union)度量,提高了网络对裂纹缺陷的识别精确度。实验结果表明,钢板表面缺陷检测算法对裂纹、夹杂、斑块、麻点、压入氧化铁皮、划痕6类缺陷的平均检测精度达到了88.9%,每秒帧数达到110.4 frame·s-1,其中小目标裂纹准确率达到75%。展开更多
Medical image segmentation has become a cornerstone for many healthcare applications,allowing for the automated extraction of critical information from images such as Computed Tomography(CT)scans,Magnetic Resonance Im...Medical image segmentation has become a cornerstone for many healthcare applications,allowing for the automated extraction of critical information from images such as Computed Tomography(CT)scans,Magnetic Resonance Imaging(MRIs),and X-rays.The introduction of U-Net in 2015 has significantly advanced segmentation capabilities,especially for small datasets commonly found in medical imaging.Since then,various modifications to the original U-Net architecture have been proposed to enhance segmentation accuracy and tackle challenges like class imbalance,data scarcity,and multi-modal image processing.This paper provides a detailed review and comparison of several U-Net-based architectures,focusing on their effectiveness in medical image segmentation tasks.We evaluate performance metrics such as Dice Similarity Coefficient(DSC)and Intersection over Union(IoU)across different U-Net variants including HmsU-Net,CrossU-Net,mResU-Net,and others.Our results indicate that architectural enhancements such as transformers,attention mechanisms,and residual connections improve segmentation performance across diverse medical imaging applications,including tumor detection,organ segmentation,and lesion identification.The study also identifies current challenges in the field,including data variability,limited dataset sizes,and issues with class imbalance.Based on these findings,the paper suggests potential future directions for improving the robustness and clinical applicability of U-Net-based models in medical image segmentation.展开更多
为降低由于电梯维保工作不到位带来的电梯安全隐患,设计一款基于机器视觉的电梯维保质量监测系统。该系统通过传感器模块检测维保人员维保场所的首次进入,进而触发维保质量监测。构建维保人员、维保工具数据集,并基于YOLOv8完成模型训练...为降低由于电梯维保工作不到位带来的电梯安全隐患,设计一款基于机器视觉的电梯维保质量监测系统。该系统通过传感器模块检测维保人员维保场所的首次进入,进而触发维保质量监测。构建维保人员、维保工具数据集,并基于YOLOv8完成模型训练,实现维保过程中的维保人员、维保工具的定位识别。进一步将维保人员边框与维保区域边框进行IOU(Intersection over Union)计算,实现区域匹配,完成维保人员在不同维保区域的维保时间计算、维保质量评价。基于Qt图形用户界面应用程序框架,构建系统的上位机可视化界面,有效避免“纸上维保”的现象,为电梯维保的质量监测技术提供依据。展开更多
为了快速精准地识别复杂果园环境下的葡萄目标,该研究基于YOLOv5s提出一种改进的葡萄检测模型(MRWYOLOv5s)。首先,为了减少模型参数量,采用轻量型网络MobileNetv3作为特征提取网络,并在MobileNetv3的bneck结构中嵌入坐标注意力模块(coor...为了快速精准地识别复杂果园环境下的葡萄目标,该研究基于YOLOv5s提出一种改进的葡萄检测模型(MRWYOLOv5s)。首先,为了减少模型参数量,采用轻量型网络MobileNetv3作为特征提取网络,并在MobileNetv3的bneck结构中嵌入坐标注意力模块(coordinate attention,CA)以加强网络的特征提取能力;其次,在颈部网络中引入RepVGG Block,融合多分支特征提升模型的检测精度,并利用RepVGG Block的结构重参数化进一步加快模型的推理速度;最后,采用基于动态非单调聚焦机制的损失(wise intersection over union loss,WIoU Loss)作为边界框回归损失函数,加速网络收敛并提高模型的检测准确率。结果表明,改进的MRW-YOLOv5s模型参数量仅为7.56 M,在测试集上的平均精度均值(mean average precision,mAP)达到97.74%,相较于原YOLOv5s模型提升了2.32个百分点,平均每幅图片的检测时间为10.03 ms,比原YOLOv5s模型减少了6.13 ms。与主流的目标检测模型SSD、RetinaNet、YOLOv4、YOLOv7和YOLOX相比,MRW-YOLOv5s模型的mAP分别高出9.89、7.53、2.12、0.91、2.42个百分点,并且在模型参数量大小和检测速度方面有着很大的优势,该研究可为果园智能化、采摘机械化提供技术支持。展开更多
文摘In positive-ion fast atom bombardment (FAB) mass spectrometry, when mono- and di- saccharides are mixed with an appropriate amount of NH4Cl, a highly abundan peak [M+NH4]+appers in FAB mass spectra . From the adduct ion [M+NH4]+, the molecular weights of mono- and di- saccharides can be determined definitively
文摘森林火点检测在林火应急救援中起着至关重要的作用.鉴于现有模型在样本质量、多尺度检测以及多视角图像泛化能力方面存在不足,以YOLOv7为基础,提出一种森林火点目标检测方法FFD-YOLO(forest fire detection based on YOLO).首先,构建多视角可见光图像森林火灾高点检测数据集FFHPV(forest fire of high point view),旨在增强模型对多视角火点知识的学习能力;其次,引入全维动态卷积,构建空间金字塔池化层(OD-SPP),以此提升模型针对多视角数据的火点特征提取能力;最后,引入具有动态非单调聚焦机制的边界框定位损失函数Wise-IoU(wise intersection over union),降低低质量数据对模型精度的影响,提高小目标火点的检测能力.实验结果表明:所提出的FFD-YOLO方法相较于YOLOv7,精度提高3.9%,召回率提高3.7%,均值平均精度提高4.0%,F1分数提高0.038;同时,在与YOLOv5、YOLOv8、DDQ(dense distinct query)、DINO(detection transformer with improved denoising anchor boxes)、Faster R-CNN、Sparse R-CNN、Mask R-CNN、FCOS和YOLOX的对比实验中,FFD-YOLO具有最高的精度75.3%、召回率73.8%、均值平均精度77.6%和F1分数0.745,验证了该方法的可行性与有效性.
基金Supported by a grant from Japan China Sasakawa Medical Fellowship。
文摘Ultrasound plays an important role not only in preoperative diagnosis but also in intraoperative guidance for liver surgery.Intraoperative ultrasound(IOUS)has become an indispensable tool for modern liver surgeons,especially for minimally invasive surgeries,partially substituting for the surgeon’s hands.In fundamental mode,Doppler mode,contrast enhancement,elastography,and real-time virtual sonography,IOUS can provide additional real-time information regarding the intrahepatic anatomy,tumor site and characteristics,macrovascular invasion,resection margin,transection plane,perfusion and outflow of the remnant liver,and local ablation efficacy for both open and minimally invasive liver resections.Identification and localization of intrahepatic lesions and surrounding structures are crucial for performing liver resection,preserving the adjacent vital vascular and bile ducts,and sparing the functional liver parenchyma.Intraoperative ultrasound can provide critical information for intraoperative decision-making and navigation.Therefore,all liver surgeons must master IOUS techniques,and IOUS should be included in the training of modern liver surgeons.Further investigation of the potential benefits and advances in these techniques will increase the use of IOUS in modern liver surgeries worldwide.This study comprehensively reviews the current use of IOUS in modern liver surgeries.
文摘为满足多数工业场景下钢板表面缺陷检测的需求,针对钢板表面缺陷检测准确率低及小目标缺陷检测率低等问题,文中提出了一种基于改进YOLOv5(You Only Look Once version 5)的钢板表面缺陷检测算法。在YOLOv5的基础上将CBAM(Convolution Block Attention Module)注意力模块嵌入到主干网络中,提高网络检测精度。加入上下文增强模块,提高了算法对小目标的检测性能。使用NWD(Normalized Wasserstein Distance)度量标准代替原YOLOv5中的IoU(Intersection over Union)度量,提高了网络对裂纹缺陷的识别精确度。实验结果表明,钢板表面缺陷检测算法对裂纹、夹杂、斑块、麻点、压入氧化铁皮、划痕6类缺陷的平均检测精度达到了88.9%,每秒帧数达到110.4 frame·s-1,其中小目标裂纹准确率达到75%。
文摘Medical image segmentation has become a cornerstone for many healthcare applications,allowing for the automated extraction of critical information from images such as Computed Tomography(CT)scans,Magnetic Resonance Imaging(MRIs),and X-rays.The introduction of U-Net in 2015 has significantly advanced segmentation capabilities,especially for small datasets commonly found in medical imaging.Since then,various modifications to the original U-Net architecture have been proposed to enhance segmentation accuracy and tackle challenges like class imbalance,data scarcity,and multi-modal image processing.This paper provides a detailed review and comparison of several U-Net-based architectures,focusing on their effectiveness in medical image segmentation tasks.We evaluate performance metrics such as Dice Similarity Coefficient(DSC)and Intersection over Union(IoU)across different U-Net variants including HmsU-Net,CrossU-Net,mResU-Net,and others.Our results indicate that architectural enhancements such as transformers,attention mechanisms,and residual connections improve segmentation performance across diverse medical imaging applications,including tumor detection,organ segmentation,and lesion identification.The study also identifies current challenges in the field,including data variability,limited dataset sizes,and issues with class imbalance.Based on these findings,the paper suggests potential future directions for improving the robustness and clinical applicability of U-Net-based models in medical image segmentation.
文摘为降低由于电梯维保工作不到位带来的电梯安全隐患,设计一款基于机器视觉的电梯维保质量监测系统。该系统通过传感器模块检测维保人员维保场所的首次进入,进而触发维保质量监测。构建维保人员、维保工具数据集,并基于YOLOv8完成模型训练,实现维保过程中的维保人员、维保工具的定位识别。进一步将维保人员边框与维保区域边框进行IOU(Intersection over Union)计算,实现区域匹配,完成维保人员在不同维保区域的维保时间计算、维保质量评价。基于Qt图形用户界面应用程序框架,构建系统的上位机可视化界面,有效避免“纸上维保”的现象,为电梯维保的质量监测技术提供依据。
文摘为了快速精准地识别复杂果园环境下的葡萄目标,该研究基于YOLOv5s提出一种改进的葡萄检测模型(MRWYOLOv5s)。首先,为了减少模型参数量,采用轻量型网络MobileNetv3作为特征提取网络,并在MobileNetv3的bneck结构中嵌入坐标注意力模块(coordinate attention,CA)以加强网络的特征提取能力;其次,在颈部网络中引入RepVGG Block,融合多分支特征提升模型的检测精度,并利用RepVGG Block的结构重参数化进一步加快模型的推理速度;最后,采用基于动态非单调聚焦机制的损失(wise intersection over union loss,WIoU Loss)作为边界框回归损失函数,加速网络收敛并提高模型的检测准确率。结果表明,改进的MRW-YOLOv5s模型参数量仅为7.56 M,在测试集上的平均精度均值(mean average precision,mAP)达到97.74%,相较于原YOLOv5s模型提升了2.32个百分点,平均每幅图片的检测时间为10.03 ms,比原YOLOv5s模型减少了6.13 ms。与主流的目标检测模型SSD、RetinaNet、YOLOv4、YOLOv7和YOLOX相比,MRW-YOLOv5s模型的mAP分别高出9.89、7.53、2.12、0.91、2.42个百分点,并且在模型参数量大小和检测速度方面有着很大的优势,该研究可为果园智能化、采摘机械化提供技术支持。
