基于计算机视觉技术的结构表面裂缝检测方法研究

计算机视觉技术用于混凝土结构表面裂缝检测,具有现场检测方便、效率高、客观性强的特点,但图像数据分析是该技术的核心,其中裂缝提取与定量测量较为复杂。为提高裂缝图像处理效率和准确率,将深度学习和数字图像处理技术相结合,提出一种裂缝检测方法。建立基于深度卷积神经网络的裂缝识别模型,在图像上自动定位裂缝并结合图像局域阈值分割方法提取裂缝。在裂缝宽度定量测量方面,采用双边滤波算法和三段线性变换对裂缝图像进行预处理,提高了裂缝边缘识别的精确度。通过改进边缘梯度法,实现裂缝最大宽度的定位和裂缝最大宽度的自动获取。该研究为全自动识别裂缝图像及高精度测量裂缝宽度提供了一种解决方法。     

基于瑞利波法对混凝土表面裂缝检测的数值研究

基于瑞利波及衍射纵波的传播时间差法,利用有限元法分析了瑞利波在含不同长度及角度裂纹的混凝土介质中的传播规律,提出了混凝土介质中表面裂缝长度和角度的计算公式,并验证了数值模拟结果的准确性。研究结果表明:在传播过程中,瑞利波一部分沿着介质表面(含裂纹表面)传播,另一部分在裂缝尖端发生模式转换,形成衍射横波和衍射纵波,弹性波传播到介质边界时被粘弹性人工边界全部吸收,消除了反射波的影响;表面裂缝长度和角度的计算结果误差基本在5%以下,只有个别数据误差较大,在一定范围内,随着裂缝长度和角度值的增大,计算结果的准确度越高。     

基于面波法的混凝土坝体裂缝深度检测研究

介绍了面波法检测裂缝深度的原理,阐述了基于面波法的混凝土坝体裂缝深度检测方案,指出该方法基本不受裂缝内部充填物的影响,检测结果较声波法更加精确、可靠。     

Image quality evaluation method for surface crack detection based on standard test chart

The use of automated equipment for surface crack detection based on digital image acquisition is becoming increasingly popular in the inspection industry. While researchers typically focus on improving the accuracy of recognition methods, the image quality is essential to the effectiveness of the algorithm. However, evaluating the quality of crack images has received little attention in computer-aided civil and infrastructure engineering. A prominent issue is whether surface cracks are visible and measurable in images. This study proposes an image quality evaluation method using an original standard test chart that simulates cracks of different widths and directions. Geometric transformations and preprocessing techniques are employed in a full-reference strategy to process the acquired crack images. The resulting information provides quantitative scores for crack visibility and measurability. The proposed Crack Structural Similarity Index is more in line with human perception and offers an accurate evaluation of real image quality. The study shows that Gaussian blur disturbance and random noise disturbance primarily affect measurability and visibility, respectively. Furthermore, the study finds that the quality of the crack image improves with increasing sensor pixel size and using a prime lens over a zoom or long zoom lens. This approach enables comparing image quality collected by different devices in the field environment and provides guidance for optimizing field acquisition parameters. In the future, the results of this study can be applied to facilitate the application of automated testing equipment and improve overall performance.     

Self-training approach for crack detection using synthesized crack images based on conditional generative adversarial network

Urban infrastructure plays a crucial role in determining the quality of life for citizens. However, given the increasing number of aging infrastructures, regular inspections are essential to prevent accidents. Deep learning studies have been conducted to detect structural damage and ensure high accuracy and reliability of these inspections. However, these detection algorithms often face challenges due to scarcity of damage data. To overcome this issue, this paper proposes a method for synthesizing crack images and utilizing them for crack detection. Initially, crack images are synthesized from labeled images by using a conditional generative adversarial network. Subsequently, a new self-training method is implemented wherein the synthesized crack images from the prediction images were incorporated into the learning process to enhance data diversity. The proposed approach yields promising results with a mean intersection over union of 80.34% and F1-score of 76.31% on average. The proposed method can aid further research on virtual image generation for crack detection, seeking to reduce the reliance on extensive image collection.     

基于冲击回波法的混凝土裂缝检测试验研究

在混凝土裂缝检测中,裂缝深度的检测精度受到检测距离的影响。为进一步研究裂缝深度与检测距离的具体关系,基于冲击回波法对裂缝进行不同检测间距的测量,并将所得到的波形信号进行傅里叶变换转化为频谱图,通过裂缝缺陷频谱计算出裂缝深度。发现当检测间距为裂缝深度的1. 5～2倍时,裂缝检测结果较为精确。     

A new test method for automatic fire detection devices

A new test method is proposed in which all automatic fire detectors, regardless of sensor type, would be subjected to a series of the same test fires. Note: Mr. Bright presented this paper at the 80th Annual Meeting of the National Fire Protection Association in Houston, Texas, May 1976.     

基于电压电荷法的静电放电能量检测方法

传统的静电放电能量检测方法多依赖于经验公式,方法简单,计算误差较大。根据电压电荷法计算静电放电能量的原理,设计并搭建了实验系统,开发了基于MATLAB APP Designer的静电放电能量计算软件,并结合静电火灾事故调查实例验证其的科学性,为静电事故的预防与调查提供技术手段。     

基于电压电荷法的静电放电能量检测方法

传统的静电放电能量检测方法多依赖于经验公式,方法简单,计算误差较大.根据电压电荷法计算静电放电能量的原理,设计并搭建了实验系统,开发了基于MATLAB APP De-signer的静电放电能量计算软件,并结合静电火灾事故调查实例验证其的科学性,为静电事故的预防与调查提供技术手段.     

基于Faster R-CNN的交通标志检测方法

文章提出了一种基于Faster R-CNN的交通标志检测方法。首先,通过采集公路两侧交通标志的图像建立数据集;其次,使用预训练的VGG16对Faster R-CNN模型的相应参数进行初始化;最后,对Faster R-CNN模型进行训练。实验表明,其mAP可达81.21%,可以满足实际需求。     

基于温度-应力试验的混凝土抗裂性仿真分析方法

温度-应力试验能够较好地评价混凝土抗裂性,但因其耗用时间长、试验成本大而制约温度-应力试验法的发展。本试验应用温度-应力试验法得到线膨胀系数、自生体积变形、徐变等变形参数随成熟度发展的函数关系,并用B4Cast温度应力分析软件模拟实际浇筑过程中混凝土的温度发展历程,通过仿真计算,对比实际工况下不同配合比混凝土的抗裂性。结果表明:应用B4Cast仿真计算软件可以得到不同温控措施下混凝土内部的温度历程,采用与实际情况接近的温度历程进行温度应力分析,才能客观评价混凝土的抗裂性;在温度应力试验结果的基础上,通过拟合得到混凝土的热膨胀系数、收缩变形、弹性模量、抗拉强度等参数随成熟度的变化规律,再将其用于B4Cast计算出的温度历程,可以对混凝土的开裂风险进行仿真分析;应用此方法得出在同等级下低热水泥的混凝土抗裂性高于中热水泥的混凝土。     

无人飞机在广西桥梁检测中的应用发展

针对目前传统的公路桥梁检测手段的局限性,创新性地引入无人机检测理念,将无人机技术、计算机技术、航空摄影技术、互联网技术结合起来,提出一种全新的公路桥梁检测思路,该方法可广泛应用于广西公路桥梁检测中,有效提高了广西公路桥梁检测效率、缩短检测周期、节省工程检测费用,并可向全区工程检测其他领域拓展。     

基于DIC技术和无网格法的裂尖应变场分析方法

针对数字图像相关(DIC)技术难以获得精准的裂纹尖端应变场问题,提出一种基于DIC技术和伽辽金无网格法(EFG)的混合方法来分析裂尖应变场。该方法基于DIC技术导出场节点建立无网格模型,采用EFG法结合衍射法表示裂纹不连续性,进而计算全场应变值。以Q235qE钢材制成的紧凑拉伸(CT)试样为例开展了准静态循环拉伸试验,通过比较裂尖应变解析解和试验结果,证明该方法的有效性和稳定性。进一步研究DIC-EFG法在裂尖塑性区域和应力强度因子(SIF)计算等方面的应用。研究结果表明:DIC-EFG混合方法建模简单,不需要荷载和约束条件,能够获得准确的包括裂尖在内的全平面应变场。相较于传统DIC技术,DIC-EFG混合方法解决疲劳裂纹附近结果缺失的问题,结合J积分法计算SIF结果更为精准,可应用于疲劳裂纹的塑性区域和SIF计算等断裂性能研究。     

基于黏度特征气体探测识别方法实验研究

提出一种基于差压原理的气体组分检测方法,通过测量待测气体流经特制黏度传感部件后产生的差压信号响应,换算获得气体黏度数据,根据泊肃叶定律及混合气体黏度理论定量关系反演获取待测气体组成。通过实验验证了利用黏度差异进行气体识别的理论和应用可行性。通过开展基于流量基准值和测量斜率的未知气体识别实验,证明两种方法均能够对4种未知气体进行精准识别。     

基于小波分析法的GPS周跳探测

在GPS测量中,如何对载波相位观测值中的周跳进行探测和修复,是精密定位数据处理中的一个十分重要的问题,概述了周跳的产生、定义及其来源,介绍了不同周跳检测量的构成,重点讨论了周跳检测的小波分析方法.     

基于压电波动法的混凝土损伤检测

混凝土作为最基本的建筑材料,在土木工程中的地位是举足轻重的。然而随着建筑结构的老化和自然灾害的作用,大多数建筑结构内部或多或少都有损伤,而这种损伤一般需要借助仪器才能被发现,因而使得人们无法及时对损伤部位做出响应。本文基于压电波动法,使用两个压电智能骨料来检测混凝土的健康状况,通过对比健康状态和损伤状态下信号的变化,并基于小波包分析的方法来得到小波包能量指数与混凝土健康状况之间的关系。     

Pipe crack identification based on the energy method and committee of neural networks

A crack identification method using an equivalent bending stiffness for cracked beam and committee of neural networks is presented. The equivalent bending stiffness is constructed based on an energy method for a straight thin-walled pipe, which has a through-the-thickness crack, subjected to bending. Several numerical analysis for a steel cantilever pipe using the equivalent bending stiffness are carried out to extract the natural frequencies and mode shapes of the cracked beam. The extracted modal properties are used in constructing a training patterns of a neural network. The input to the neural network consists of the modal properties and the output is composed of the crack location and size. Multiple neural networks are constructed and each individual network is trained independently with the different initial synaptic weights. Then, the estimated crack locations and sizes from different neural networks are averaged. Crack detection is carried out for 16 damage cases using the proposed method, and the identified crack locations and sizes agree reasonably well with the exact values.     

基于Griffith强度理论的岩石裂纹起裂经验预测方法研究

 裂纹起裂强度是岩石破坏过程中的重要应力阈值,研究岩石起裂准则对于揭示其破坏机制及预测围岩工程性质有着重要意义。首先进行青砂岩试样的单轴及三轴压缩起裂试验,并基于多种应变响应分析其中的起裂机制及细观破坏特征,指出局部张拉应力集中是起裂破坏的主因,总结提出低围压条件下的张开型起裂模型及高围压条件下的滑动型起裂模型。然后基于Griffith强度理论分析压应力场中岩石缺陷端部的局部最大张拉应力,其大小随差应力 的升高而增大,同时在围压条件下受表面摩擦作用的影响较大。针对岩石细观起裂机制提出起裂预测经验准则,准则中引入起裂参数 作为围压影响系数以表征摩擦作用,从而适用于不同围压条件下的起裂破坏预测。利用3组起裂试验结果对经验准则进行验证,其准确性及实用性明显优于传统线性起裂准则。最后通过分析不同围压下岩石起裂强度与峰值强度之比 ,发现试样在围压60 MPa以下时其起裂破坏属于细观张拉破坏机制。     

基于岩石峰前起裂及峰后特征的脆性评价方法

现有研究表明岩石脆性与其内部微裂纹起裂、扩展有密切联系,但考虑岩石起裂特征的脆性评价指数并不多见,在对脆性理论以及起裂特征分析的基础上,提出了一种基于岩石峰前起裂及峰后特征的脆性评价方法.首先,结合George和Tarasov&Potvin等对脆性的描述和定义,从理论上分析了基于岩石峰前起裂及峰后特征脆性评价方法的可行...     

Signal detection aspects in automatic fire detection

Automatic fire detectors can be considered as signal detectors for a special “signal-in-noise” detection problem. The paper uses the classical model for the measurable signal x(t) = s(t) + n(t) and gives an analysis of the signal component s(t), which is generated by the fire and which is to be detected. Using a general matched filter concept in combination with rather general properties of the fire signal component s(t), several forms of fire signal detectors can be developed which are optimum detectors in a rather general sense. These detectors are given as simple signal processing algorithms which can be realized by using a microprocessor or a special integrated electronic circuit. The efficiency of fire detectors depends very much on the behavior of the noise component n(t) which in the practical case of heat or smoke detection must be considered as highly self-correlated, in other words n(t) varies slowly with time compared with the fire signal component s(t). In this case it is nearly impossible to calculate the detector efficiency in advance for signal detection situations which may be interesting for practical cases.

In order to study the behavior and the efficiency of these detectors a Monte Carlo method has been used, and a measuring procedure has been established with suitable accuracy and in a way which covers the practical need.