Intelligent Detection Model Based on a Fully Convolutional Neural Network for Pavement Cracks

The crack is a common pavement failure problem.A lack of periodic maintenance will result in extending the cracks and damage the pavement,which will affect the normal use of the road.Therefore,it is significant to establish an efficient intelligent identification model for pavement cracks.The neural network is a method of simulating animal nervous systems using gradient descent to predict results by learning a weight matrix.It has been widely used in geotechnical engineering,computer vision,medicine,and other fields.However,there are three major problems in the application of neural networks to crack identification.There are too few layers,extracted crack features are not complete,and the method lacks the efficiency to calculate the whole picture.In this study,a fully convolutional neural network based on ResNet-101 is used to establish an intelligent identification model of pavement crack regions.This method,using a convolutional layer instead of a fully connected layer,realizes full convolution and accelerates calculation.The region proposals come from the feature map at the end of the base network,which avoids multiple computations of the same picture.Online hard example mining and data-augmentation techniques are adopted to improve the model’s recognition accuracy.We trained and tested Concrete Crack Images for Classification(CCIC),which is a public dataset collected using smartphones,and the Crack Image Database(CIDB),which was automatically collected using vehicle-mounted charge-coupled device cameras,with identification accuracy reaching 91.4%and 86.4%,respectively.The proposed model has a higher recognition accuracy and recall rate than Faster RCNN and different depth models,and can extract more complete and accurate crack features in CIDB.We also analyzed translation processing,fuzzy,scaling,and distorted images.The proposed model shows a strong robustness and stability,and can automatically identify image cracks of different forms.It has broad application prospects in practical engineering problems.

New Lite YOLOv4-Tiny Algorithm and Application on Crack Intelligent Detection

Conforming to the rapidly increasing market demand of crack detection for tall buildings,the idea of integrating deep network technology into wall-climbing robot for crack detection is put forward in this paper.Taking the dependence and hardware requirements when deployed on such edge devices as Raspberry Pi into consideration,the Darknet neural network is selected as the basic framework for detection.In order to improve the inference efficiency on edge devices and avoid the possible premature over-fitting of deep networks,the lite YOLOv4-tiny algorithm is then improved from the original YOLOv4-tiny algorithm and its structure is illustrated using Netron accordingly.The images downloaded from Internet and taken from the buildings in campus are processed to form crack detection data sets,which are trained on personal computer with the AlexeyAB version of Darknet to generate weight files.Meanwhile,the AlexeyAB version of Darknet accelerated by NNpack package is deployed on Raspberry Pi 4B,and the crack detection experiments are carried out.Some characteristics,e.g.,fast speed and lower false detection rate of the lite YOLOv4-tiny algorithm,are confirmed by comparison with those of original YOLOv4-tiny algorithm.The innovations of this paper focus on the simple network structure,fewer network layers,and earlier forward transmission of features to prevent over-fitting,showing the new lite neural network exceeds the original YOLOv4-tiny network significantly.

Creation of Extension Knowledge Base System About Intelligent Detection in Dendrobium Huoshanense Photosynthesis Process

Aiming at the limitations of the existing knowledge representations in intelligent detection,a novel extension-based knowledge representation(EKR) is proposed.The definitions,grammar rules,and formal semantics of EKR are presented.A rhombus solving strategy(RSS) based on EKR is discussed in detail,including creation of the problem oriented model,extension operator,the solution formation of contradictions problem and extended inference of matter-element.A knowledge base system based on EKR and RSS is developed,which is applied in intelligent detection in the Dendrobium huoshanense photosynthesis process(DHPP).More reasonable results are obtained than traditional rule-based system.The EKR is feasible in intelligent detection to solve the limitations of traditional knowledge representations.

Intelligent Ammunition Detection and Classification System Using Convolutional Neural Network

Security is a significant issue for everyone due to new and creative ways to commit cybercrime.The Closed-Circuit Television(CCTV)systems are being installed in offices,houses,shopping malls,and on streets to protect lives.Operators monitor CCTV;however,it is difficult for a single person to monitor the actions of multiple people at one time.Consequently,there is a dire need for an automated monitoring system that detects a person with ammunition or any other harmful material Based on our research and findings of this study,we have designed a new Intelligent Ammunition Detection and Classification(IADC)system using Convolutional Neural Network(CNN).The proposed system is designed to identify persons carrying weapons and ammunition using CCTV cameras.When weapons are identified,the cameras sound an alarm.In the proposed IADC system,CNN was used to detect firearms and ammunition.The CNN model which is a Deep Learning technique consists of neural networks,most commonly applied to analyzing visual imagery has gained popularity for unstructured(images,videos)data classification.Additionally,this system generates an early warning through detection of ammunition before conditions become critical.Hence the faster and earlier the prediction,the lower the response time,loses and potential victims.The proposed IADC system provides better results than earlier published models like VGGNet,OverFeat-1,OverFeat-2,and OverFeat-3.

Intelligent Intrusion Detection System Model Using Rough Neural Network

A model of intelligent intrusion detection based on rough neural network (RNN), which combines the neural network and rough set, is presented. It works by capturing network packets to identify network intrusions or malicious attacks using RNN with sub-nets. The sub-net is constructed by detection-oriented signatures extracted using rough set theory to detect different intrusions. It is proved that RNN detection method has the merits of adaptive, high universality, high convergence speed, easy upgrading and management.

Intelligent Spectrum Detection Model Based on Compressed Sensing in Cognitive Radio Network

In view of the uncertainty of the status of primary users in cognitive networks and the fact that the random detection strategy cannot guarantee cognitive users to accurately find available channels,this paper proposes a joint random detection strategy using the idle cognitive users in cognitive wireless networks.After adding idle cognitive users for detection,the compressed sensing model is employed to describe the number of available channels obtained by the cognitive base station to derive the detection performance of the cognitive network at this time.Both theoretical analysis and simulation results show that using idle cognitive users can reduce service delay and improve the throughput of cognitive networks.After considering the time occupied by cognitive users to report detection information,the optimal participation number of idle cognitive users in joint detection is obtained through the optimization algorithm.

Intelligent identifi cation method for near-surface ground fi ssures based on seismic data

Taking a study area in Jinzhong Basin in Qixian County,Shanxi Province,as an example,this work performs an intelligent interpretation of ground fissures.On the basis of a complete analysis of the regional geological background in the study area,dip-steering cube operation and median filtering of seismic data were performed using fast Fourier transform to improve the continuity of seismic events and eliminate random noise.A total of 200 stratigraphic continuous sample training points and 500 discontinuous training points were obtained from the processed seismic data.Thereafter,a variety of attributes(coherence,curvature,amplitude,frequency,etc.)were extracted as the input for the multilayer perceptron neural network training.During the training period,the training results were traced by normalized root mean square error(RMSE)and misclassifi cation.The training results showed a downward trend during the training period.The misclassifi cation curve was stable at 0.3,and the normalized RMSE curve was stable at 0.68.When the value of the normalized RMSE curve reached the minimum,the training was terminated,and the training results were extended to the whole data volume to obtain the attribute cube of intelligent ground fi ssure detection.The characteristics of ground fi ssures were analyzed and identifi ed from the sections and slices.A total of 11 ground fissures were finally interpreted.The interpretation results showed that the dip angles were 60°-85°,the fault throws were 0-43 m,and the extension lengths were 300-1,100 m in the whole area.The strike of 73%of the ground fi ssures was consistent with the direction of the regional tectonic settings.Specifi cally,four ground fi ssures coincided with the surface disclosed,and the verifi cation rate reached 100%.In conclusion,the intelligent ground fi ssure detection attribute based on the dip-steering cube is eff ective in predicting the spatial distribution of ground fi ssures.

Semi-automatic Video Annotation Tool to Generate Ground Truth for Intelligent Video Surveillance Systems

Generating ground truth data for developing object detection algorithms of intelligent surveillance systems is a considerably important yet time-consuming task; therefore, a user-friendly tool to annotate videos efficiently and accurately is required. In this paper, the development of a semi-automatic video annotation tool is described. For efficiency, the developed tool can automatically generate the initial annotation data for the input videos utilizing automatic object detection modules, which are developed independently and registered in the tool. To guarantee the accuracy of the ground truth data, the system also has several user-friendly functions to help users check and edit the initial annotation data generated by the automatic object detection modules. According to the experiment＇s results, employing the developed annotation tool is considerably beneficial for reducing annotation time; when compared to manual annotation schemes, using the tool resulted in an annotation time reduction of up to 2.3 times.

Artificial intelligence in polyp detection-where are we and where are we headed?

The goal of artificial intelligence in colonoscopy is to improve adenoma detection rate and reduce interval colorectal cancer.Artificial intelligence in polyp detection during colonoscopy has evolved tremendously over the last decade mainly due to the implementation of neural networks.Computer aided detection(CADe)utilizing neural networks allows real time detection of polyps and adenomas.Current CADe systems are built in single centers by multidisciplinary teams and have only been utilized in limited clinical research studies.We review the most recent prospective randomized controlled trials here.These randomized control trials,both non-blinded and blinded,demonstrated increase in adenoma and polyp detection rates when endoscopists used CADe systems vs standard high definition colonoscopes.Increase of polyps and adenomas detected were mainly small and sessile in nature.CADe systems were found to be safe with little added time to the overall procedure.Results are promising as more CADe have shown to have ability to increase accuracy and improve quality of colonoscopy.Overall limitations included selection bias as all trials built and utilized different CADe developed at their own institutions,non-blinded arms,and question of external validity.

Track Defects Recognition Based on Axle-Box Vibration Acceleration and Deep- Learning Techniques

As an important component of load transfer,various fatigue damages occur in the track as the rail service life and train traffic increase gradually,such as rail corrugation,rail joint damage,uneven thermite welds,rail squats fas-tener defects,etc.Real-time recognition of track defects plays a vital role in ensuring the safe and stable operation of rail transit.In this paper,an intelligent and innovative method is proposed to detect the track defects by using axle-box vibration acceleration and deep learning network,and the coexistence of the above-mentioned typical track defects in the track system is considered.Firstly,the dynamic relationship between the track defects(using the example of the fastening defects)and the axle-box vibration acceleration(ABVA)is investigated using the dynamic vehicle-track model.Then,a simulation model for the coupled dynamics of the vehicle and track with different track defects is established,and the wavelet power spectrum(WPS)analysis is performed for the vibra-tion acceleration signals of the axle box to extract the characteristic response.Lastly,using wavelet spectrum photos as input,an automatic detection technique based on the deep convolution neural network(DCNN)is sug-gested to realize the real-time intelligent detection and identification of various track problems.Thefindings demonstrate that the suggested approach achieves a 96.72%classification accuracy.

超薄电子玻璃冲击检测试验机设计与开发

随着智能产品的不断发展,对超薄电子玻璃的需求越来越多,性能要求也越来越高,对其各项性能检测也就显得越来越重要。智能产品朝着手持式、小型化的方向发展,其往往会发生跌落、冲击等破坏现象,抗冲击强度能直接反映、评价或者判断超薄电子玻璃的抵抗冲击能力,是反映超薄电子玻璃应用性的重要指标,因此对超薄电子玻璃的抗冲击检测显得尤其重要。设计与开发了一种新型结构的超薄电子玻璃冲击检测试验机,以适应新标准的冲击强度检测要求,并达到智能化检测的新目标,大幅提高检测效率,对新标准的落地、实施及促进超薄电子玻璃产品质量提升贡献一份力量。

火试金自动化检测系统的研制及应用

传统火试金分析一直沿用手动作业模式,劳动强度大、效率低、职业健康安全风险高。研制的火试金自动化检测系统由自动化配料系统、自动化混料添加系统、自动化熔样系统、自动化灰吹系统及自动化分金系统组成,对应配料、混料、熔融、灰吹及分金5个分析步骤。该系统实现了火试金的自动化检测,检测结果准确可靠,与人工检测结果无差异,且操作过程统一稳定,不受人员主观因素影响。投产应用后,大幅提高了检测效率,比人工模式提高了167百分点以上,降低了人员劳动强度,保障了人员的职业健康安全。火试金自动化检测系统具有较高的推广应用价值。

非接触式氨纶架检测装置

为了在积极送纱设备上使氨纶纱达到更好效果,改善布面平整度及相应弹性,提高产品质量,开发一款新型氨纶送纱装置。其工作原理是通过机械杠杆进行低张力纱线检测及控制,调节不同配重档位控制氨纶纱线送纱张力及效果,另外新的光学检测位置调整使调节更加简单方便。该装置使用新型非接触式光学检测设计,通过加工工艺的改进,使输出张力波动更小,性能也更加稳定,可应用于各种大圆机、小圆机等设备上。

基于云边端的空间目标探测感知系统架构设计

空间目标探测感知系统作为未来信息化作战的关键要素,为现代化战争提供了重要的天域预警信息支撑,文中面对未来空间目标呈现小型化、星座化、异变化等特点,将云边端架构引入空间目标探测感知系统的架构设计中。首先,分析了目前空间目标探测感知系统面临的问题和系统需求;然后,结合云边端架构的理念提出了一种空间目标探测感知系统架构的设计方法,考虑当前探测感知装备现状合理设计了云中心、边中心和端节点的系统组成,结合空间目标探测流程设计了“两环、一库、四要素”的系统运行逻辑架构;最后,总结了该架构具备的先进能力特征。

选煤厂灰分快速检测技术发展现状及展望

研究灰分快速检测技术的发展现状对于研制符合智能选煤厂建设需要的灰分测定技术具有重要意义,系统地梳理了快速灰化法、有源及无源在线测试和国标法自动测灰技术,指出现有技术存在的主要问题并提出建议。研究发现,我国矿区型选煤厂整体技术水平相对滞后,不同区域或企业在采制化过程中的自动化和准确程度等方面还存在较大差距。快速灰化法存在效率低、工人劳动强度大、样品代表性差等问题,已无法适应选煤厂要求;低能γ射线反散射,高能γ射线湮没辐射、双能γ射线透射、中子活化、X射线荧光光谱和激光诱导击穿光谱(LIBS)等有源在线测试方法,其测试精度受煤种、粒度、灰分组成及含量、水分等煤质特性参数以及选煤工艺条件影响较大,且还存在设备投资成本、放射源管理和日常维护管理成本相对较高等问题;国标法全自动测灰技术实现了采样、制样和测灰的自动化运行,测试精度高,但测试时间相对较长。而选煤厂灰分快速检测技术研发应按照快速测定、高精度、高准确度、无人化、标准化的整体要求,根据选煤生产的实际需求和选煤工艺进行定制化研发,通过数字孪生等技术提高智能化和数字化水平。

基于深度学习的炉内钢坯关键点检测方法

在部分生产线上,钢坯从炉内出炉的过程中,需要依靠人眼判断钢坯是否到达指定位置,到位后再利用推钢机推动炉内的钢坯完成出炉过程。在这个过程中,人眼长时间观察摄像头屏幕容易疲劳,人工劳动强度大成本高,生产工作效率较低。针对以上问题,文中提出利用机器视觉系统替代人类视觉系统进行钢坯位置的实时定位,首先将钢坯定位问题转换为关键点检测问题,然后提出了基于ResNet网络和基于关键点分割网络(Key Point Segmentation Network,KPSN)的两种模型来进行关键点检测,最后,通过测试和分析所提出的两种方法,提出了多方法融合的关键点检测方案,降低了极端情况下误检的风险,实际应用表明,文中所提方法具有较高的鲁棒性,达到了实际应用的要求。

基于PCA-ELM的弹载组合导航智能故障检测算法

针对传统PCA-ELM(主元分析-极限学习机)算法分类效果稳定性差和准确率不高等问题,结合弹载组合导航系统卡尔曼滤波器,提出一种改进PCA-ELM故障检测方法。首先,分析了PCA算法负载矩阵与卡尔曼滤波新息协方差矩阵的关系,构造新的权系数矩阵,并引入极限学习机对权系数矩阵进行参数优化,将参数优化后的负载矩阵进行故障分析。最后,将该算法首次应用于弹载组合导航系统。仿真实验表明,在检测斜坡型故障方面,检测速度和检测正确率均优于传统PCA,MSS(多子集分离法)及AIME(自主完好性外推法)算法。

智能化工况模拟与检测系统研究与开发

智能化工况模拟与检测系统研究开发,研究了智能化土壤-植物-机械工况模拟与检测技术,智能化种子精量播种模拟与检测技术、机械化乳品采集检测模拟与检测技术、农产品组分与缺陷在线检测技术,开发了智能化土壤植物机械工况模拟实验系统、喷药机械田间作业工况模拟装置、配方施肥变量控制模拟装置、智能化精量播种模拟系统、机械化乳品采集检测模拟系统、农产品组分和缺陷声光检测模拟系统等。突破了高性能先进装备共性和关键技术,为新技术发展提供平台,缩短新技术、新产品、新装备的研究开发成果转化为生产力的周期,为国内研发、中试、生产等关键环节提供专业化试验基地。全天候智能化土壤-植物-机械工况模拟实验系统,采用坑式直线型土槽及轨道进行农机具或部分整机的性能试验,实现了被测机具或整机的耕深、动力输出轴转速和牵引速度的自动控制,并能够对土壤坚实度、土壤含水率、动力输出轴转速和扭矩、牵引速度等10余项数据进行实时采集和处理。采用集成化设计,集土壤恢复系统、机具悬挂系统、测试系统及控制系统于一体,开发了六分力门架测力系统,创造性地采用了＂智能化工况模拟试验＂这一全新的试验模式,进行该试验台控制系统的软硬件设计,提高了试验效率,减轻了试验人员的劳动强度,缩短了试验及研究周期。全天候智能化精量播种模拟系统主要用于条播排种器和精密（单粒）排种器的排种性能试验,检测出排种器的各项性能指标,可在胶带速度为1.8~12 km/h情况下进行试验。模拟系统能自动控制排种器转轴的工作速度、胶带的运行速度及风机压力;计算机视觉系统实时监测排种器的工作过程,并利用图像检测技术自动精确测得排种器所排种子粒距、粒数等参数,从而计算处理国标试验要求的合格指数、重播指数和漏播指数等排种性能指标,以及播种精度指标（平均值、标准差、变异系数）,并报表打印及输出频率直方图,可存储原始播种录像。机械化乳品采集检测模拟系统在国内首次研制成功挤奶系统脉动性能、挤奶量自动计量装置性能、真空设备性能在线检测与遥测、自动脱杯系统提升力测试等4种乳品采集性能模拟试验与检测技术装置与设备。农产品组分和缺陷声光检测模拟系统建立了适合西瓜内部品质无损检测的光源装置和光谱采集系统、西瓜声学特性检测试验台、基于声波信号衰减率的西瓜糖度检测方法。

智能交通运动检测模型与计算方法的研究

介绍了一种精确识别违章车辆的数学模型,它借助运动检测理论计算方法,配合虚拟线圈,充分考虑光照及空气抖动的影响,利用二次检测算法,能够较准确的识别车辆.在武威的智能指挥中心的实际应用表明该算法很好的解决了违章车辆的识别问题,对城市智能化发展中车辆的违章识别具有很重要的意义.

基于集成技术的三维焊缝检测传感器设计

设计了一种基于集成技术的三维焊缝检测传感器。为解决焊接加工自动化过程中三维焊缝的跟踪检测,将GMR焊枪悬浮高度传感器、焊缝检测传感器以及相应的微处理器有机结合并集成,从而使传感器实现小型化、智能化。针对焊接过程的非线性、多变性以及易受焊疤等干扰源干扰等问题,采用了智能控制策略实现三维焊缝检测,减小焊缝检测的盲区范围。实验结果表明:该传感器能够检测出空间三维焊缝,且焊缝跟踪精度高,达到了设计要求。