Application of Boron-doped Diamond Film Electrode in Fast Detection of Samonella in Water

[Objective] The study aimed to discuss the application of boron-doped diamond （BDD） film electrode in fast detection of samonella in water. [ Method] Boron-doped diamond film electrode was prepared and used as the working electrode in fast detection of salmonella in water using chronoamberometry, and the oxidation mechanism of the electrode acting on salmonella was discussed. [ Result] Compared with traditional biologi- cal methods, chronoamperometry could detect the number of salmonellae in water more simply, rapidly and sensitively. [ Conclusion] The method of using BDD electrode to detect salmonella quantity will be widely applied in future.

Research on fast detection method of infrared small targets under resourceconstrained conditions

Infrared small target detection is a common task in infrared image processing.Under limited computa⁃tional resources.Traditional methods for infrared small target detection face a trade-off between the detection rate and the accuracy.A fast infrared small target detection method tailored for resource-constrained conditions is pro⁃posed for the YOLOv5s model.This method introduces an additional small target detection head and replaces the original Intersection over Union(IoU)metric with Normalized Wasserstein Distance(NWD),while considering both the detection accuracy and the detection speed of infrared small targets.Experimental results demonstrate that the proposed algorithm achieves a maximum effective detection speed of 95 FPS on a 15 W TPU,while reach⁃ing a maximum effective detection accuracy of 91.9 AP@0.5,effectively improving the efficiency of infrared small target detection under resource-constrained conditions.

Fast detection algorithm for cracks on tunnel linings based on deep semantic segmentation

An algorithm based on deep semantic segmentation called LC-DeepLab is proposed for detecting the trends and geometries of cracks on tunnel linings at the pixel level.The proposed method addresses the low accuracy of tunnel crack segmentation and the slow detection speed of conventional models in complex backgrounds.The novel algorithm is based on the DeepLabv3+network framework.A lighter backbone network was used for feature extraction.Next,an efficient shallow feature fusion module that extracts crack features across pixels is designed to improve the edges of crack segmentation.Finally,an efficient attention module that significantly improves the anti-interference ability of the model in complex backgrounds is validated.Four classic semantic segmentation algorithms(fully convolutional network,pyramid scene parsing network,U-Net,and DeepLabv3+)are selected for comparative analysis to verify the effectiveness of the proposed algorithm.The experimental results show that LC-DeepLab can accurately segment and highlight cracks from tunnel linings in complex backgrounds,and the accuracy(mean intersection over union)is 78.26%.The LC-DeepLab can achieve a real-time segmentation of 416×416×3 defect images with 46.98 f/s and 21.85 Mb parameters.

基于灰度均值的自适应FAST角点检测优化算法

光照不均、突变引起的灰度变化会影响图像特征检测效果。为此,设计一种基于灰度均值的自适应FAST-9-12角点检测算法。首先,利用特征点的延展性设计一种小面积双重检测模板,减少像素点与中心点的比较次数,提高区域正检率和检测速度;其次,依据图像局部灰度均值,在每个像素点检测模板内自适应调整阈值,避免灰度变化影响检测效果;最后,根据柔性非极大值抑制的思想设计角点半径抑制原则,筛选鲁棒性更强的角点。在Inria遥感影像数据集上的实验结果表明,FAST-9-12角点检测速度比FAST-12-16,FAST-9-16两种模板提高22%左右,因自适应阈值的提取方式不易受光照影响,检测准确率分别提高4.16和3.11个百分点,实现了图像特征快速和精准检测。

基于FAST算法的核动力旋转机械故障诊断

基于FAST算法设计了核动力旋转机械故障诊断方法。提取候选区故障特征,在卷积层内输出候选区向量,通过计算图像尺寸得到候选区内故障特征函数值;基于FAST算法标记旋转机械故障点,对最大池化值与最小池化值进行计算,得到标准层的梯度收敛情况;设计了核动力旋转机械故障诊断算法,得到旋转机械故障诊断方法。对比4种不同旋转机械故障诊断方法的实验结果可知,针对6种不同的故障特征,基于FAST算法的核动力旋转机械故障诊断方法的对数似然率均为最大,诊断准确率更高。

基于改进FAST的电力工程图纸识别与校核技术

针对人工评审电力施工图预算工程量过程中存在的耗时费力及准确性较低等问题,设计了一种基于改进FAST算法与VGG网络的工程图纸识别与数据校核模型。该模型针对传统FAST算法误判率高、抗噪声性能差等不足,提出了基于积分核的改进FAST算法,以降低角点检测误判率和检测时间。并采用VGG网络进行输入特征与图纸数据校核之间的关联分析,从而实现了电力工程图纸符号智能识别与电力工程量的智能校核。算例验证结果表明,所提IFAST-VGG算法对图纸识别与工程量校核的平均准确率分别为90.6%和91.6%,均高于其他对比算法,可以应用于实际工程中。

RVP Fast与Seeplex PneumoBacter ACE Detection assay联合诊断呼吸道感染病原体

目的采用RVP Fast与Seeplex PneumoBacter ACE Detection assay两种方法,联合诊断呼吸道感染病原体。方法收集南京市呼吸道感染患者咽拭子标本67份,提取RNA及DNA,呼吸道病毒群快速检测(Respiratory Virus Panel Fast Array,RVP Fast)技术检测呼吸道病毒,Seeplex PneumoBacter ACE Detection assay检测呼吸道六重病原体。结果 67份标本中,呼吸道病毒阳性仅4例,主要为肠鼻病毒。呼吸道六重病原体检测中,阳性标本为26例,其中肺炎链球菌、流感嗜血杆菌、肺炎衣原体阳性较多。结论 RVP Fast与Seeplex PneumoBacter ACE Detection assay联合使用,有利于明确呼吸道感染病原体。

基于改进Fast-SCNN的裂缝图像实时分割算法

裂缝检测是一项关键工程任务,针对现有的主流裂缝语义分割模型参数量大、计算量高、实时性弱等问题,提出一种基于Fast-SCNN(fast segmentation convolution neural network)改进的裂缝图像实时分割算法。首先,该算法在Fast-SCNN基础上优化了空间金字塔池化模块SPP(spatial pyramid pooling)存在像素位置信息丢失以及计算量大的不足,提出了一种轻量级的特征金字塔注意力模块;其次,改进了上采样的方式,充分考虑像素之间的关系,提出了一种轻量级的位置自注意力模块用于上采样,以此来提升检测精度;最后,双分支的各自输出通过注意力门突显裂缝相关区域和抑制无关背景。所提算法能够为模型提供更为精确的像素级别的注意力,更加有效识别细小裂缝和提升复杂背景裂缝分割的鲁棒性。实验结果表明:与现有的主流模型和其他轻量级模型相比,该算法进一步平衡了分割精度与检测速度,在裂缝数据集上达到80.31%的平均交并比,F1 score为76.74%,参数量为1.20 M,计算量不足1 G,推理速度达到151 f/s,对裂缝图像实时分割检测任务具有较高的应用价值。

Fast Rate Fault Detection Filter for Multirate Sampled-data Systems

This paper focuses on the fast rate fault detection filter (FDF) problem for a class of multirate sampled-data (MSD) systems. A lifting technique is used to convert such an MSD system into a linear time-invariant discrete-time one and an unknown input observer (UIO) is considered as FDF to generate residual. The design of FDF is formulated as an H∞ optimization problem and a solvable condition as well as an optimal solution are derived. The causality of the residual generator can be guaranteed so that the fast rate residual can be implemented via inverse lifting. A numerical example is included to demonstrate the feasibility of the obtained results.

Fast Electrical Detection of Carcinoembryonic Antigen Based on AlGaN/GaN High Electron Mobility Transistor Aptasensor

As one of the most important tumor-associated antigens of colorectal adenocarcinoma, the carcinoembryonic antigen （CEA） threatens human health seriously ali over the globe. Fast electrical and highly sensitive detection of the CEA with A1GaN/GaN high electron mobility transistor is demonstrated experimentally. To achieve a low detection limit, the Au-gated sensing area of the sensor is functionalized with a CEA aptamer instead of the corresponding antibody. The proposed aptasensor has successfully detected different concentrations （ranging from 50picogram/milliliter （pg/ml） to 50 nanogram/milliliter （ng/ml）） of CEA and achieved a detection limit as low as 50pg/ml at Vas = 0.5 V. The drain-source current shows a c/ear increase of 11.5μA under this bias.

Unit panel node detection by CNN on FAST reflector

The Five-hundred-meter Aperture Spherical radio Telescope(FAST) has an active reflector.During observations, the reflector will be deformed into a paraboloid 300 meters in diameter. To improve its surface accuracy, we propose a scheme for photogrammetry to measure the positions of 2226 nodes on the reflector. The way to detect the nodes in the photos is the key problem in this application of photogrammetry. This paper applies a convolutional neural network(CNN) with candidate regions to detect the nodes in the photos. Experimental results show a high recognition rate of 91.5%, which is much higher than the recognition rate for traditional edge detection.

Experiments on mine-detection with fast neutron activation

Fast neutron activation of nitrogen and oxygen contained in the explosives used for simulated mine samples has been preliminarily carried out in our laboratory. By spectroscopic analysis of characteristic γ-rays emitted from activated nitrogen and oxygen, mine can be identified almost instantly. This technique integrated with robottes would be a method for mine scavenging.

Enhancing the Accuracy for Target Detection by Using Fast Orthogonal Search (FOS)

Target detection and bearing estimation are mainly obtained through spectral analysis of received signals. The detection performance of the periodogram and its variants methods is evaluated. The variants methods Performance evaluation through the Receiver Operating Characteristics （ROCs） are presented and compared from the viewpoint of probability of detection （Pd）, probability of false alarm （Pfa） by computer simulation. When the sinusoid frequency does not correspond to one of the spectral bins （mid-bin frequency situation）, the performance of all the mentioned detectors degrades. This research investigates the development of a bearing estimation method using Fast Orthogonal Search （FOS） to enhance spectral estimation which, improves both target detection and bearing estimation in case of low SNR inputs.

涂片与培养方法检测痰液和肺泡灌洗液中抗酸菌诊断肺结核的管理研究

目的评价痰液和肺泡灌洗液联合抗酸菌检测的临床应用价值,为肺结核临床诊疗决策提供参考。方法以在同一周内并行痰液涂片染色查找抗酸菌和/或分枝杆菌培养检查的可疑肺结核患者为研究对象,从医院LIS系统中收集、整理2021年1月—7月的相关检测数据,分析4种检测结果的同一性与差异性。结果①在272例患者中,痰涂片、肺泡灌洗液涂片、痰培养、肺泡灌洗液培养(以下分别称为“A”“B”“C”“D”)的阳性检出率分别为14.71%(40/272)、19.49%(53/272)、25.00%(67/268)和31.90%(74/232),4种检测作为平行试验的阳性率为37.50%(102/272);②A^(+)C^(+)、A^(-)C^(+)、A^(+)C^(-)、A^(-)C^(-)、A^(-)C N(上标中的“+”“-”“N”分别代表“阳性”“阴性”“未检测”,下同)和B^(+)D^(+)、B^(-)D^(+)、B^(+)D^(-)、B^(-)D^(-)、B^(-)D^(N)结果组合分别占14.71%(40/272)、13.97%(38/272)、0、69.85%(190/272)、1.47%(4/272)和19.12%(52/272)、8.82%(24/272)、0.37%(1/272)、56.99%(155/272)、14.71%(40/272);③A^(+)B^(+)、A^(+)B^(-)、A^(-)B^(+)、A^(-)B^(-)和A^(+)D^(+)、A^(+)D^(-)、A^(+)D^(N)、A^(-)D^(+)、A^(-)D^(-)、A^(-)D^(N)、A^(N)D^(+)、A^(N)D^(-)、A^(N)D^(N)结果组合分别占14.71%(40/272)、0、4.78%(13/272)、80.51%(219/272)和19.12%(52/272)、5.51%(15/272)、4.04(11/272)、8.09%(22/272)、51.74%(140/272)、10.29%(28/272)、0.74%(2/272)、0.37%(1/272)、0.37%(1/272)。结论较通常应用痰液检测抗酸菌而言,行肺泡灌洗液抗酸菌检测可以大幅提高肺结核病原学确诊率,值得临床大力推广应用。

基于HigherHRNet的煤矿井下人体姿态估计快速网络研究

煤矿井下人体姿态的快速估计是井下作业智慧安全检测的重要前提。为解决煤矿井下多尘多雾、照明不足及颜色相融等问题,提高人体姿态估计关键点分配准确度及网络运行速度,研究新的Optimising HigherHRNet(OH-HRNet)快速网络模型:对HigherHRNet模型的轻量化设计、关键点分配进行深入研究,提出了基于注意力机制的记忆卷积模块及强化骨骼约束的关键点分配算法,并改进了算法的损失函数。在煤矿井下场景数据集和COCO公开数据集上的实验结果表明:OH-HRNet在GPU的速度上是LitePose的1.06倍,平均精度均值mAP提高了7.4%,平均召回率均值mAR提高了14.0%,可以实现更有效的智慧安全检测。

基于视觉图像的飞行区大型目标监视方法改进研究

机场飞行区现使用的场面监视方法存在着定位偏差较大、不稳定、易跳变、皆为点源定位等问题。针对这些问题,设计了基于视觉图像的飞行区监视方法,实现快速准确的目标检测和轮廓定位,使飞行区监视更加稳定精确。提出了一种基于MobileNetV3和YOLOv5的网络模型(以下称为MobileNetV3-YOLOv5),即在YOLOv5的主干中使用MobileNetV3,来提高对目标的检测速度和准确度;提出了一种基于优化特征点提取的改进定向快速旋转简报(Oriented FAST and Rotated BRIEF,ORB)算法,将图像分割成多个区域,分别提取每个区域的特征点,从而提高目标识别框内区域的特征点识别数量,再进行特征点聚类筛选,最后根据识别目标类型采用最小包围盒进行轮廓划分,得到目标的轮廓定位。试验结果表明:MobileNetV3-YOLOv5方法对比原始YOLOv5模型,在识别目标准确率方面提升5百分点,在效率方面提升14张/s;同时在0~60 m的范围内,轮廓估计误差仅为2.9%;体现了所提出的监视方法的有效性,可以提升飞行区监视定位准确性和运行安全性。

基于点击-去点击化学反应的Hg^(2+)荧光探针检测体系构建

基于杂化罗丹明与巯基化合物之间的点击化学反应及巯基化合物与金属汞离子的配位反应导致的去点击化学反应机理,构建香豆素杂化罗丹明探针-1,3,5-苯三硫酚-Hg^(2+)的三元荧光检测体系。在该体系中,探针能够与1,3,5-苯三硫酚中的巯基基团发生加成反应,破坏探针的共轭结构,从而导致探针的红色荧光猝灭。加入Hg^(2+)后,Hg^(2+)与1,3,5-苯三硫酚通过配位作用结合,使其从探针结构上离去,探针的共轭结构及红色荧光恢复。体系的荧光强度与Hg^(2+)在5×10^(-9)~1×10^(-8)、1×10^(-8)~1.1×10^(-7)mol/L浓度范围内呈良好的线性关系,检测限为2×10^(-9)mol/L,等倍量的常见重金属离子对Hg^(2+)的特异性识别无干扰。应用于实际饮用水样品中Hg^(2+)含量检测,回收率可达到86.7%~110.2%。

基于DN-YOLOv5遥感目标快速检测方法

针对遥感目标检测中检测速度慢,网络计算量大的问题,提出一种基于DN-YOLOv5的遥感目标快速检测方法。目标检测锚框采用K-means++重新聚类,以减少网络的计算量;修改算法主干网络部分的CSP1_X模块以减少参数量,并在主干网络进行双密集连接,加强卷积操作对特征的利用率;将算法模型中的CBL模块的Leak ReLU激活函数替换成H-swish激活函数,以提高网络检测速度。将快速目标检测方法 DN-YOLOv5在公开的DIOR遥感数据集上进行验证。结果表明,改进后的快速检测方法,模型大小为48 MB,降低了约47.2%,推理时延为33.8 ms,推理加速了约39%。

基于快速时频图的海面无人机目标多特征检测

海面无人机(UAV)的检测属于海杂波背景下的小目标检测问题,多特征联合检测是解决此类问题的有效途径。针对已有的时频三特征检测方法在特征提取阶段计算复杂度过大、难以实现实时检测的问题,提出了一种基于快速时频图的海面无人机多特征检测方法。首先,对雷达复回波数据进行分段快速傅里叶变换,将计算得到的多普勒幅度谱沿多普勒维对齐拼接从而构建快速时频图;其次,对快速时频图进行归一化,达到杂波抑制和增强目标回波的目的,并基于归一化的快速时频图提取三种时频特征;然后,利用快速凸包学习算法训练给定虚警概率下的检测判决区域;最后,通过实测UAV数据验证并分析了所提方法的有效性。

一种快速鲁棒的LOG-FAST角点算法

基于高时间效率的FAST算法,提出了一种快速鲁棒的FAST-LOG角点算法。使用直方图均衡化方法对图像进行增强,提高图像成分的清晰度并消除图像中光照强度的影响;运用拉普拉斯-高斯函数对图像进行卷积,实现图像的高斯平滑和增强边缘,及对噪声最大化的抑制;最后使用FAST算子检测角点。对比实验证明,新算子对于添加高斯噪声的分辨率为640*480的图像,其检测时间可达到0.05s;对光照不同的图像具有相近的检测性能;角点重复率可达98%。该算子可应用于实时视频图像的处理,为开发基于视觉的实时智能车辆预警系统提供了新的研究思路。