Cracking the Snake Detection Theory: The Subcortical Visual Pathway as a Major Player in Cultural Transformations

According to the proposed hypothesis, graphic characters trigger the subcortical visual route. The reaction discussed is very weak. Yet, its very existence has an unusual importance: characters and (occluded) venomous snakeskin patterns reveal themselves as conflatable. Furthermore, following tractogra- phic research, a functional segregation of the subcortical pathway is to be presupposed. Thus, there can’t be a later dissociation of two stimuli previously associated. The outcomes of lecture will gradually appear probabilistically (much) more peaceful than encountering a venomous snake, though, and thus a continuous lessening of the reaction is expectable. Here, on one hand, it is relevant that the subcortical visual pathway goes to the amygdala. The reactions we describe tap into goal-oriented processes, and they will do that unfettered. On the other hand, in the case of characters, since the beginning, fear has been converted into appetition to a great degree. This process should be fostered in the presence of light. In this way, luminosity might become a conditioned stimulus for attraction. In this case, a Pavlovian addiction for light will foster, yet also—from the point of view of reward feeling—counterbalance the lessening of the stimulation elicited by characters. The addiction we refer to is one towards light accompanied by graphic signs. Yet, as opposed to the case of the luminous medium, the attention captured by the later ones taken for themselves is continuously reduced.

Simultaneous determination of three curcuminoids in Curcuma longa L.by high performance liquid chromatography coupled with electrochemical detection

A novel method for analysis of three active components curcumin, demethoxycurcumin and bisdemethoxycurcumin in Curcuma longa L. was developed by HPLC coupled with electrochemical detection. Three curcuminoids were well separated on a C18 column and detected with high sensitivity. A mobile phase containing acetonitrile and 10 mM Na2HPO4-H3PO4 （pH 5.0） （50：50, v/v） was used. Good linearity was obtained in the range of 0.208-41.6, 0.197-39.4, and 0.227-114μM for curcumin, demethoxycurcumin and bisdemethoxycurcumin respectively. The limit of detection reached up to 10 ？ 8 M, which was lower than that by UV detection. The relative standard deviations （RSDs） ranged from 1.06%to 1.88%for intra-day precision and from 4.30%to 5.79%for inter-day precision, respectively. The proposed method has been applied in real herb sample and recoveries ranging from 86.3%to 111%were obtained.

Semi-Supervised Learning Based Big Data-Driven Anomaly Detection in Mobile Wireless Networks

With rising capacity demand in mobile networks, the infrastructure is also becoming increasingly denser and complex. This results in collection of larger amount of raw data(big data) that is generated at different levels of network architecture and is typically underutilized. To unleash its full value, innovative machine learning algorithms need to be utilized in order to extract valuable insights which can be used for improving the overall network's performance. Additionally, a major challenge for network operators is to cope up with increasing number of complete(or partial) cell outages and to simultaneously reduce operational expenditure. This paper contributes towards the aforementioned problems by exploiting big data generated from the core network of 4 G LTE-A to detect network's anomalous behavior. We present a semi-supervised statistical-based anomaly detection technique to identify in time: first, unusually low user activity region depicting sleeping cell, which is a special case of cell outage; and second, unusually high user traffic area corresponding to a situation where special action such as additional resource allocation, fault avoidance solution etc. may be needed. Achieved results demonstrate that the proposed method can be used for timely and reliable anomaly detection in current and future cellular networks.

The interaction between a screw dislocation and a wedge-shaped crack in one-dimensional hexagonal piezoelectric quasicrystals

Based on the fundamental equations of piezoelasticity of quasicrystal material,we investigated the interaction between a screw dislocation and a wedge-shaped crack in the piezoelectricity of one-dimensional hexagonal quasicrystals.Explicit analytical solutions are obtained for stress and electric displacement intensity factors of the crack,as well as the force on dislocation.The derivation is based on the conformal mapping method and the perturbation technique.The influences of the wedge angle and dislocation location on the image force are also discussed.The results obtained in this paper can be fully reduced to some special cases already available or deriving new ones.

Feasibility of Trace Alcohol Congener Detection and Identification Using Laser-Induced Breakdown Spectroscopy

In this paper, a feasible scheme is reported for the detection and identification of trace alcohol congeners that have identical elemental composition using laser-induced breakdown spectroscopy （LIBS）. In the scheme, an intensive pulsed laser is used to break down trace alcohol samples and the optical emission spectra of the induced plasma are collected for the detection and identifq ication of alcohol molecules. In order to prepare trace alcohol samples, pure ethanol or methanol is bubbled by argon carrier gas and then mixed into matrix gases. The key issue for the scheme is to constitute indices from the LIBS data of the alcohol samples. Two indices are found to be suitable for alcohol detection and identification. One is the emission intensity ratio （denoted as H/C） of the hydrogen line （653.3 nm） to the carbon line （247.9 nm） for identification and the other is the ratio of the carbon line （as C/Ar） or the hydrogen line （as H/Ar） to the argon lines （866.7 nm） for quantitative detection. The calibration experiment result shows that the index H/C is specific for alcohol congeners while almost being independent of alcohol concentration. In detail, the H/C keeps a specific constant of 34 and 23 respectively for ethanol and methanol. In the meanwhile, the C/Ar and H/Ar indices respond almost linearly to the alcohol concentration below 1300 ppm, and are therefore competent for concentration measurement. With the indices, trace alcohol concentration measurement achieves a limit of 140 ppm using a laser pulse energy of 300 mJ.

Stresses due to an adhesion crack in T-shaped junction of two orthotropic plates

The general solution of stresses is derived for a T-shaped junction of two thin plates with an adhesion crack. The plates are orthotropic. A shear force is applied on the crack surface. The analysis is based on the supposition that the stresses in each plate can be approximated by a plane stress condition. The results obtained are verified by numerical calculation of FEM.

Filtering algorithm of line structured light for long-distance obstacle detection

Since unmanned ground vehicles often encounter concave and convex obstacles in wild ground, a filtering algorithm using line structured light to detect these long distance obstacles is proposed. For the line structured light image, a ranked-order based adaptively extremum median （RAEM） filter algorithm on salt and pepper noise is presented. In the algorithm, firstly effective points and noise points in a filtering window are differentiated; then the gray values of noise points are replaced by the medium of gray values of the effective pixels, with the efficient points＇ gray values unchanged; in the end this algorithm is proved to be efficient by experiments. Experimental resuits demonstrate that the image blur, resulting into proposed algorithm can remove noise points effectively and minimize the protecting the edge information as much as possible.

基于双层聚类信息的极光亚暴自动检测

极光亚暴与太阳风和地球磁场的耦合过程有着紧密的联系,对其发生和发展机制的研究,有助于深入地分析行星际磁场、地球磁层和地球电离层的相互作用,了解太阳风携带的大量能量在地球空间的输运过程,对地球空间环境预警具有重要的意义。Polar卫星搭载的紫外极光图像成像仪能够全天候地获取紫外极光图像,在紫外极光图像中可以完整地展示出极区极光的亮度和尺度变化,尤其是可以清晰地展示出亚暴膨胀相的极光点亮和亮斑膨胀现象。现有的极光亚暴事件检测方法通常需要人工设计特征和相关规则库,耗时耗力。本文利用紫外极光图像提出了基于双层聚类信息的亚暴事件检测方法,实现了紫外极光图像数据中的亚暴事件自动检测。同时,针对极光亚暴事件检测依赖手工设计特征,设计了子空间聚类指导的三维卷积特征自动提取网络;针对极光图像帧间存在成像角度差异,利用地磁纬度和磁地方时信息对极光序列中图像的空间位置进行校正;针对卫星成像位置变化导致的成像噪声,利用极光图像级聚类保留极光亮斑区域和剔除未成像或噪声区域。主观和客观实验结果表明,本算法提升了亚暴事件检测的查全率。

基于改进YOLOv5s的道路裂缝检测算法

为了解决道路巡检系统光学传感器采集的裂缝图像中颜色特征不明显且尺寸不规则造成检测精度不高、泛化能力不足的问题,提出改进YOLOv5s的裂缝检测算法。将结合深度可分离卷积(Depthwise Separable Convolution, DSC)的全局注意力(Global Attention Mechanism, GAM)引入主干特征提取网络,在降低注意力复杂度的同时获得丰富的跨维度特征,增强了裂缝的识别能力;采用空间金字塔软池化网络(Spatial Pyramid Softpool, SPSF),通过Softpool池化保留多维语义以减少信息弥散,提高了边界框回归的准确性;在颈部特征增强网络,运用空洞深度可分离卷积(Atrous DSC)进行下采样,通过扩大感受野加强深层和浅层信息的聚合能力,提高裂缝识别的泛化性。经过在自制道路裂缝数据集上的实验,相较于YOLOv5s,改进算法的mAP提高2.2%,有效提升了道路裂缝检测的准确性和对不同背景下裂缝识别的泛化能力。

低空复杂风场全天候雷达精细探测技术

低空复杂风场通常指距离地面高度600 m以下,局地风速、风向快速变化的风场.其全天候大范围精细探测是雷达科学、气象学等领域的长期难题.低空复杂风场的全天候雷达精细探测涉及传感器设计、三维风场反演和全天候融合等关键技术.本文从激光雷达探测、微波雷达探测、全天候融合及特征危害提取等方面对风场探测的技术内涵及研究现状进行了梳理,并分析了发展趋势,为低空复杂风场全天候精细探测技术发展与应用提供参考.

对深度学习中目标定位不确定度评定的讨论

得益于深度学习技术的快速发展,高准确率的目标自动定位得以实现,这为各领域的智能化转型提供了极大的助力。然而,尽管大多数情况下基于深度学习的目标定位方法都能够获得较为精确的输出,但一些误差较大的定位结果还是难以避免。正因为缺少规范的不确定度评定,所以那些误差较大的定位结果难以被有效地消除,进而影响了基于深度学习的目标定位方法的工程化应用。从符合计量规范的测量不确定度入手,讨论了当前深度学习中目标定位技术评价指标的意义和不足,并提出了对规范化目标定位技术不确定度评定的建议。

基于多视投影的电力杆塔点云绝缘子串快速提取

在无人机自动巡检中,需要根据电力走廊带点云中绝缘子串的位置进行航线规划。由于电力走廊带长度跨越较大,杆塔数量众多,人工进行绝缘子串位置标注的工作量巨大。同时点云数据散乱无序,在三维空间中直接定位绝缘子串耗时长。为此,本文提出多视二维投影,快速提取三维杆塔点云中的绝缘子串。该方法首先利用鲁棒主成分分析,在俯视图中分析杆塔顶端横担水平方向,统一杆塔点云的朝向。接着,在侧视图中通过霍夫直线检测,对电力线高度进行快速定位,并根据电力线与横担的空间关系,将相应的绝缘子串分为顶层悬挂、竖直绝缘子串与水平绝缘子串三类。最后,在多视投影中进行角点检测,对绝缘子串与电力线、杆塔的连接端点进行定位,实现快速的绝缘子串提取。本文提出的方法在50 km的电力走廊带中的97座不同类型杆塔的867个绝缘子串上进行了提取实验,证明了该方法能对不同类型绝缘子串进行快速提取。

一种基于光学的水表面微波跨介质探测方法

水下声信号的传播和探测研究主要集中在散射、反射和传播损失等方面,而对其在水面引起的波动关注较少。本文在理论上推导了在水下平面波入射在水-空气界面激励的水表面微波与声源强度和频率的关系。这种水表面微波的振幅在纳米级别,振动频率和水下平面波的频率相同,在水-空气界面的扩散距离则随频率的升高而强烈衰减。基于水表面微波的特性,结合激光探测的工作原理,设计了一套跨介质水下目标检测系统。并为该系统可能受到的低频海浪噪声干扰和反射干扰提供了应对方式。仿真结果表明:该系统分辨率高,具有隐蔽性、易于维护性、无接触性等特点,对低频声源的检测效果较好。该方法结合了声波在水中和光波在空气中传播各自的优点,有望成为水下目标检测的一种新手段。

基于华为鸿蒙的智能路灯控制系统设计

针对传统路灯控制方式缺乏灵活性的问题,文章设计了一种以华为海思Hi3861芯片作为控制核心智能路灯控制系统。本系统通过光照强度传感器检测路灯周围光照强度;根据声音传感器检测路灯周围声音信息,判断是否有车辆经过;利用红外传感器检测行人路过情况。系统搭载华为鸿蒙操作系统,通过MQTT协议将采集到的数据传送到华为云平台,对路灯进行控制,从而降低电能的浪费。实验表明,本系统具有通信稳定可靠、灵活性高、网络支持完善等特点。

深度学习技术在地震储层预测中的应用及挑战

传统地震储层预测技术已无法满足储层精细评价的需求,深度学习具有强大的特征提取和高维数据处理能力,近年来广泛应用于地震储层预测并取得了较好的效果。为此,本文深入讨论深度学习技术在地震储层预测中的应用、进展及它在实际工作中面临的挑战,并提出未来的发展方向。主要认识有:(1)在烃类定性检测方面,深度学习技术有助于综合利用多属性地震数据去提高效率和预测结果的准确率;在定量预测方面,深度学习技术可以更精准地逼近地震数据与目标之间复杂的非线性关系,实现储层的精细定量评价。(2)深度学习技术的应用面临的挑战主要是标签数据不足和样本不均衡等容易导致模型过拟合,泛化能力差;模型复杂,计算成本高;模型的“黑匣子”特征使预测结果缺乏物理可解释性;缺乏定性预测模型的评价标准和高精度的不确定性量化算法。(3)未来的研究方向应致力于克服数据可用性的不足和深度学习的局限性等,构建地球物理知识图谱,实现多源数据与知识的有效融合、共享,将深度学习与反馈强化学习等其他机器学习算法相结合,为油气勘探和开发提供更可靠的技术支撑。

食品中黄曲霉毒素M_(1)的间接竞争酶联免疫法的建立

该文建立一种间接竞争酶联免疫法(indirect competitive enzyme-linked immunosorbent assay,ic-ELISA)快速测定食品中黄曲霉毒素M_(1)(aflatoxin M_(1),AFM_(1))的分析方法。AFM_(1)标准品用甲醇稀释,AFM_(1)标准品与AFM_(1)全抗原竞争结合AFM_(1)单克隆抗体,利用3,3′,5,5′-四甲基联苯胺(3,3′,5,5′-tetramethylbenzidine,TMB)单组分显色液显色后,酶标仪测定吸光度。在优化好的条件下,浓度范围为9.743~2.605×10^(3)pg/mL时具有良好的线性关系,相关系数(determination coefficient,R^(2))为0.9927,检出限IC_(10)为3.84 pg/mL,加标回收率为88.43%~105.75%。该方法适用性好、操作简单、灵敏度高,满足食品中AFM_(1)分析检测的需求。

New targets number estimation method under colored noise background

A multiple targets detection method based on spatial smoothing （MTDSS） is proposed to solve the problem of the source number estimation under the colored noise background. The forward and backward smoothing based on auxiliary vectors which are received data on some specific elements is computed. By the spatial smoothing with auxiliary vectors, the correlated signals are decorrelated, and the colored noise is partially alleviated. The correlation matrix formed from the cross correlations between subarray data and auxiliary vectors is computed. By exploring the second-order statistics property of the covariance matrix, a threshold based on Gerschgorin radii of the smoothing correlation matrix is set to estimate the number of sources. Simulations and experimental results validate that MTDSS has an effective performance under the condition of the colored noise background and coherent sources, and MTDSS is robust with the correlated factor of signals and noise.

线圈耦合距对坑道瞬变电磁场的影响与校正方法

水害是我国煤层开采面临的主要灾害之一,坑道瞬变电磁方法作为一种有力的坑道水害探测手段,近年来已广泛应用于我国各大煤矿区,但受坑道空间限制,线圈与围岩表面存在不等的耦合距,导致不同测点观测的瞬变场存在差异,进而影响岩层富水性的判定精度。为此,基于数值仿真技术研究了坑道内有和无金属支护条件下耦合距对瞬变场的影响,其结果表明:(1)坑道内无金属支护时,耦合距对瞬变场的影响主要集中在早期,且随瞬变场时间延迟而快速减小,对实测瞬变场数据的影响基本可以忽略。(2)坑道内存在金属支护时,瞬变场曲线存在明显的拐点,其对应时间因围岩电阻率增大而延长;与无金属支护的瞬变场相比,拐点前瞬变场响应幅值受金属影响显著增大且无明显规律;拐点后瞬变场幅值有较大变化,但衰减趋势一致;耦合距对瞬变场响应的影响大,尤其是坑道掘进正前方观测点,其对耦合距的变化更为敏感;不同探测方位不同围岩电阻率及不同耦合距条件下获得的瞬变场曲线相关系数趋向于1,曲线之间具有对数平移特性。据此提出了瞬变场晚期数据平移校正方法,并针对坑道掘进前方低阻球状模型进行了数值验证,与校正前的数据对比表明,校正后的瞬变场数据对低阻体的判定精度更高,有效降低了耦合距对瞬变场观测的影响。现场试验进一步佐证了该方法的可靠性,可为坑道水害防治提供技术支撑。

基于深度学习的近红外光谱图像处理的农作物虫害检测研究

采集农作物虫害图像时常因外界干扰使特征区域划分不明,导致虫害检测精度不足,为此,提出基于深度学习的农作物虫害近红外光谱图像分割算法。利用分块匹配模型对近红外光谱图像进行视觉纹理提取,通过高斯滤波技术去除噪声干扰。采用边缘特征重构和角点匹配方法,构建近红外光谱图像分割和相似度度量模型,采用分块帧点检测的方法对目标关键特征点定位和包络检测,多模态特征匹配后得到虫害特征分布,完成虫害检测。测试结果表明,该方法对虫害定位均方根误差平均为0.14,耗时平均为256.4 ms,检测精度平均为96.68%,具有较好的应用性能。

基于数字孪生的电梯乘客异常行为建模与识别方法

为了解决异常行为识别研究中存在的异常案例缺乏、训练样本稀缺问题,以垂直电梯的乘客异常行为监测为任务需求,提出了基于数字孪生的电梯乘客异常行为建模与识别方法。通过搭建电梯乘客行为监测数字孪生系统架构,完成了电梯运行状态与乘客行为的虚实映射。基于数字孪生场景与人体行为建模理论构建了电梯乘客异常行为案例,扩充了异常行为的孪生数据。利用改进的OpenPose获取骨骼特征,并使用PCA-DNN训练分类模型,实现了融合孪生数据的乘客异常行为快速识别。实验结果表明,该方法不仅展现了应用数字孪生技术创建异常数据的便捷性、高效性以及安全性,还验证了使用孪生数据进行模型训练的可行性、可靠性以及准确性。