• 전자공학회논문지
• /
• 제54권6호
• /
• pp.80-89
• /
• 2017

기존의 수작업으로 이루어지는 터널에서의 균열 검출은 점검자의 주관에 따라 균열을 판별하기 때문에 객관성을 보장하기 어렵다. 이러한 문제를 해결하기 위해서 터널에서 획득된 영상을 기반으로 균열을 검출하는 시스템이 많이 제안되었다. 하지만 기존의 방법은 터널 내부의 조명 상태, 균열 이외의 기타 에지 등 잡음에 상당히 민감하다. 이러한 단점은 터널의 상태에 따라 알고리즘의 성능을 크게 제한시킨다. 본 논문에서는 이러한 단점을 극복하기 위하여 컨볼루셔널 인코더-디코더 네트워크(Convolutional encoder-decoder network)를 이용한 균열 검출 방법을 제안한다. 제안하는 방법은 재현율과 정확률의 비교를 통하여 기존 연구에 비해 성능이 크게 향상되었음을 보였다.

• Smart Structures and Systems
• /
• 제32권3호
• /
• pp.135-151
• /
• 2023

Using 3D asphalt pavement surface data, a deep and multiscale network named CrackNet-M is proposed in this paper for pixel-level crack detection for improvements in both accuracy and robustness. The CrackNet-M consists of four function-specific architectural modules: a central branch net (CBN), a crack map enhancement (CME) module, three pooling feature pyramids (PFP), and an output layer. The CBN maintains crack boundaries using no pooling reductions throughout all convolutional layers. The CME applies a pooling layer to enhance potential thin cracks for better continuity, consuming no data loss and attenuation when working jointly with CBN. The PFP modules implement direct down-sampling and pyramidal up-sampling with multiscale contexts specifically for the detection of thick cracks and exclusion of non-crack patterns. Finally, the output layer is optimized with a skip layer supervision technique proposed to further improve the network performance. Compared with traditional supervisions, the skip layer supervision brings about not only significant performance gains with respect to both accuracy and robustness but a faster convergence rate. CrackNet-M was trained on a total of 2,500 pixel-wise annotated 3D pavement images and finely scaled with another 200 images with full considerations on accuracy and efficiency. CrackNet-M can potentially achieve crack detection in real-time with a processing speed of 40 ms/image. The experimental results on 500 testing images demonstrate that CrackNet-M can effectively detect both thick and thin cracks from various pavement surfaces with a high level of Precision (94.28%), Recall (93.89%), and F-measure (94.04%). In addition, the proposed CrackNet-M compares favorably to other well-developed networks with respect to the detection of thin cracks as well as the removal of shoulder drop-offs.

• 한국정보기술학회논문지
• /
• 제17권9호
• /
• pp.99-112
• /
• 2019

도로 포장면의 크랙(crack)은 도로포장 구조의 열화를 입증하는 중요한 신호와 증상이다. 카메라 영상기반 도로포장 크랙 탐지는 강도 비균질성, 위상 복잡성, 낮은 대조도 및 노이즈성의 텍스처 배경 때문에 어려운 문제이다. 본 논문은 흑백영상에 대하여 깊은 신경망(DNN)에 기반하여 픽셀수준의 도로 크랙 탐지 및 분할 문제에 대해 다룬다. 변형된 U-net 네트워크와 고수준 특징 네트워크를 포함하는 새로운 DNN 구조를 제안한다. 본 연구의 중요 기여는 융합 층을 통해 공급되는 이들 네트워크의 결합 방법이다. 우리가 아는 한, 본 연구는 보도블럭 크랙 분할 및 탐지 문제를 결합을 소개한 최초의 논문이다. 크랙 탐지 및 분할의 시스템 성능은 새로운 구조를 사용하여 급격히 향상되었다. 제안된 시스템을 2개의 공개 데이터셋­크랙 포레스트 데이터셋(CFD)와 AigleRN 데이터셋­에 대하여 구현하고 평가하였다. 본 논문의 시스템은 여덟 가지의 최신 알고리즘과 같은 데이터셋으로 실험을 하였을 때, 가장 뛰어난 결과를 보여주었다.

• 대한기계학회논문집A
• /
• 제25권3호
• /
• pp.353-361
• /
• 2001

Recently the boundary element method has been developed swiftly. The boundary element method is an efficient and accurate means for analysis of two dimensional elastic crack problems. This paper is concerned with the evaluation and the prediction of the stress intensity factor(SIF) for the crack emanating from the circular hole using boundary element method-neural network. The SIF of the crack emanating from the hole was calculated by using boundary element method. Neural network is used to evaluate and to predict SIF from the results of boundary element method. The organized neural network system (structure of four processing element) was learned with the accuracy 99%. The learned neural network system could be evaluated and predicted with the accuracy of 83.3% and 71.4% (in cases of SIF and virtual SIF). Thus the proposed boundary element method-neural network is very useful to estimate the SIF.

• 한국생산제조학회지
• /
• 제6권1호
• /
• pp.27-33
• /
• 1997

A neural network approach has been developed to determine the depth of a surface breaking crack in a steel plate from ultrasonic backscattering data. The network is trained by the use of feedforward three-layered network together with a back-scattering algorithm for error correction. The signal used for crack insonification is a mode converted 70$^{\circ}$transverse wave. A numerical analysis of back scattered field is carried out based on elastic wave theory, by the use of the boundary element method. The numerical data are calibrated by comparison with experimental data. The numerical analysis provides synthetic data for the training of the network. The training data have been calculated for cracks with specified increments of the crack depth. The performance of the network has been tested on other synthetic data and experimental data which are different from the training data.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2013년도 추계학술대회 논문집
• /
• pp.320-324
• /
• 2013

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 different initial synaptic weights. Then, the estimated crack locations and sizes from different neural networks are averaged. Experimental crack detection is carried out for 3 damage cases using the proposed method, and the identified crack locations and sizes agree reasonably well with the exact values.

• Journal of Mechanical Science and Technology
• /
• 제16권4호
• /
• pp.454-467
• /
• 2002

It has been established that a crack has an important effect on the dynamic behavior of a structure. This effect depends mainly on the location and depth of the crack. Toidentifythelocation and depth of a crack in a structure, a method is presented in this paper which uses neuro-fuzzy-evolutionary technique, that is, Adaptive-Network-based Fuzzy Inference System (ANFIS) solved via hybrid learning algorithm (the back-propagation gradient descent and the least-squares method) and Continuous Evolutionary Algorithms (CEAs) solving sir ale objective optimization problems with a continuous function and continuous search space efficiently are unified. With this ANFIS and CEAs, it is possible to formulate the inverse problem. ANFIS is used to obtain the input(the location and depth of a crack) - output(the structural Eigenfrequencies) relation of the structural system. CEAs are used to identify the crack location and depth by minimizing the difference from the measured frequencies. We have tried this new idea on beam structures and the results are promising.

• Smart Structures and Systems
• /
• 제30권2호
• /
• pp.173-181
• /
• 2022

The current paper presents a technique to detect crack in non-uniform cantilever-type pipe beams, that have step changes in the properties of their cross sections, restrained by a translational and rotational spring with a tip mass at the free end. An equation for estimating the natural frequencies for the non-uniform beams is derived using the boundary and continuity conditions, and an equivalent bending stiffness for cracked beam is applied to calculate the natural frequencies of the cracked beam. An experimental study for a step-changed non-uniform cantilever-type pipe beam restrained by bolts with a tip mass is carried out to verify the proposed method. The translational and rotational spring constants are updated using the neural network technique to the results of the experiment for intact case in order to establish a baseline model for the subsequent crack detection. Then, several numerical simulations for the specimen are carried out using the derived equation for estimating the natural frequencies of the cracked beam to construct a set of training patterns of a neural network. The crack locations and sizes are identified using the trained neural network for the 5 damage cases. It is found that the crack locations and sizes are reasonably well estimated from a practical point of view. And it is considered that the usefulness of the proposed method for structural health monitoring of the step-changed non-uniform cantilever-type pipe beam-like structures elastically restrained in the ground and have a tip mass at the free end could be verified.

• 콘크리트학회논문집
• /
• 제17권3호
• /
• pp.369-374
• /
• 2005

이 연구의 목적은 화상처리 기법과 신경회로망을 이용하여 다섯가지 균열 패턴 즉, 횡방향, 종방향, 대각선($-45^{\circ}$) 대각선($+45^{\circ}$) 그리고 비방향성 균열의 패턴을 인식할 수 있는 기법을 제안하는 것이다. 제안된 화상처리 알고리즘과 인공 신경회로망 모델은 MATLAB 언어를 이용하여 구현하였다. 인공 신경회로망의 입력층에 들어갈 패턴인자는 Total projection technique를 통해 구하였으며, 인공 신경회로망의 구조(은닉층의 수와 은닉노드의 수)와 가중치 값은 가상 균열 화상을 사용하여 학습을 통해 결정하였다. 인공 신경회로망의 학습은 Bayesian regularization 기법을 도입함으로써 과적합 문제가 발생하지 않도록 하였으며, 이 연구에서 제안한 기법의 적합성을 판정하기 위하여 총 38개의 실제 균열 화상을 사용하여 시험하였다. 검증 시험 결과내에서는 이 연구에서 제안한 기법이 사람의 균열 패턴 인식결과와 정확히 일치하는 결과것으로 나타났다.

• 한국정밀공학회지
• /
• 제16권11호
• /
• pp.158-165
• /
• 1999

It has been established that a crack has an important effect on the dynamic behavior of a structure. This effect depends mainly on the location and depth of the crack. To identify the location and depth of a crack in a structure, a method is presented in this paper which uses hybrid neuro-genetic technique. Feed-forward multilayer neural networks trained by back-propagation are used to learn the input)the location and dept of a crack)-output(the structural eigenfrequencies) relation of the structural system. With this neural network and genetic algorithm, it is possible to formulate the inverse problem. Neural network training algorithm is the back propagation algorithm with the momentum method to attain stable convergence in the training process and with the adaptive learning rate method to speed up convergence. Finally, genetic algorithm is used to fine the minimum square error.