• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.25 no.6
• /
• pp.399-405
• /
• 2015

In this work we use the mode decomposition technique employing chirplet transform, which is able to separate the individual modes from dispersive and multimodal waveform measured with the magnetostrictive sensor. The mode decomposition technique is also used to estimate the time-frequency centers and individual energies of the reflection, which would be used to locate and characterize axial defects. The arrival times of the separated modes are calculated and the axial defect lengths can be evaluated by using the estimated arrival time. Results from an experiment on a carbon steel pipe are presented and it is shown that the accurate and quantitative defect characterization could become enabled using the proposed technique.

• International conference on construction engineering and project management
• /
• 2009.05a
• /
• pp.449-456
• /
• 2009

Image processing has been applied to steel bridge defect recognition since 1990s. Compare to human visual inspection, image processing provides a more objective and accurate way of assessment. Since shade and shadow may sometimes occur when taking bridge coating images, non-uniform illumination problems should be considered. By means of color image processing, this paper aims to mitigate the illumination effect for bridge coating assessment. Furthermore, the adaptable ellipse method (AEM) is proposed to recognize mild rust colors. Finally, AEM will be compared to the K-Means algorithm, a popular recognition method, to show its advantage.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.124-125
• /
• 2023

Defect detection during liquid crystal display (LCD) manufacturing has always been a critical challenge. This study aims to address this issue by proposing a data augmentation method based on generative adversarial networks (GAN) to improve defect identification accuracy in LCD production. By leveraging synthetically generated image data from GAN, we effectively augment the original dataset to make it more representative and diverse. This data augmentation strategy enhances the model's generalization capability and robustness on real-world data. Compared to traditional data augmentation techniques, the synthetic data from GAN are more realistic, diverse and broadly distributed. Experimental results demonstrate that training models with GAN-generated data combined with the original dataset significantly improves the detection accuracy of critical defects in LCD manufacturing, compared to using the original dataset alone. This study provides an effective data augmentation approach for intelligent quality control in LCD production.

• Journal of the Korea Institute of Building Construction
• /
• v.20 no.4
• /
• pp.357-365
• /
• 2020

While numbers of apartment housing are continually rising in the domestic housing construction industry, apartment contractors are currently developing plane models, upgrading facilities, and relevant technology, and investing much efforts to meet the higher demands of consumers. However, the construction process of apartment housing involves the intricate properties of the construction industry such as materials, workforce, equipment, weather, and unpredictable situations. If any of these factors becomes discordant and results in interference and interruption of the construction process, then defects, both functional and aesthetic, are likely to occur due to errors in the plan of industry organizers and constructors. Therefore, this research identifies the types of defects in an apartment construction project and analyzes their relative importance. Firstly, this research reviews the previous research trends and will reduce the needs of this research. Afterward, defect repair costs corresponding to the different defect types are calculated by applying results of the research and performing frequency analysis on defect types included in 'Tenant preliminary research' on apartments constructed by Company A. As a result of analyzing the importance of defect type, the top six activities, including tile, floor, paper hanging, PL window, cabinetry, and kitchen cabinet, are found to be of high importance, and the top six activities in question need of repair and management of defects first. The results of this study will help establish a plan to initially respond to such problems as refusal to move in and filing a defect suit against delay in repairing defects.

• Korean Journal of Construction Engineering and Management
• /
• v.6 no.3 s.25
• /
• pp.128-134
• /
• 2005

Up to now, the various water proofing methods and materials have been developed. For the water proofing methods, poly-urethane membrane method is one of the commonly used and increase market share in water proofing industry due to it's many advantages. However, in spite of it's many advantages over other water proofing methods, water proofing defect occurs frequently. With this respect, the study investigate water proofing defect causes in roof water proofing. The study investigate 13 water proofing construction site in Sunchon city in oder to find urethane membrane defection causes and their type. As a results of the study, the followings are founded. 1) Among various water proofing defection causes, problems of water remain phenomenon due to surface horizontal level defect which occupy $25\;\%$ of total defect causes is the most commonly occur. 2) The second defect cause which occupy $15\;\%$ of total defect causes is the swell up phenomenon due to surface dry problem. For the prevention of water proofing defection in roof using urethane membrane, the followings are recommended. 1) Faultless surface treatment before using urethane membrane 2) Develop improved urethane membrane material 3) Improve urethane membrane construction technique

• Journal of the Korean Society for Nondestructive Testing
• /
• v.30 no.4
• /
• pp.302-310
• /
• 2010

In this paper, we studied Bagging neural network for predicting defect size of steam generator(SG) tube in nuclear power plant. Bagging is a method for creating an ensemble of estimator based on bootstrap sampling. For predicting defect size of SG tube, we first generated eddy current testing signals for 4 defect patterns of SG tube with various widths and depths. Then, we constructed single neural network(SNN) and Bagging neural network(BNN) to estimate width and depth of each defect. The estimation performance of SNN and BNN were measured by means of peak error. According to our experiment result, average peak error of SNN and BNN for estimating defect depth were 0.117 and 0.089mm, respectively. Also, in the case of estimating defect width, average peak error of SNN and BNN were 0.494 and 0.306mm, respectively. This shows that the estimation performance of BNN is superior to that of SNN.

• Korean Journal of Construction Engineering and Management
• /
• v.6 no.1 s.23
• /
• pp.140-150
• /
• 2005

Recently, quality inspection and defect management have been considered as one of the major issues for increasing customer satisfaction and corporate image in domestic construction industry. However, the quality inspection and defect management have not been performed systematically because of insufficient field managers, the excessive amount of documents, complicated work process and difficulty in communicating construction information. Therefore, the field manager could not perform the quality inspection and defect management work in time as well as the reliability of recorded quality and defect data was decreased. The primary objective of this study is to propose a quality inspection and defect management system using information technology which enables field managers to efficiently gather the information of defection in apartment housing. It is anticipated that the effective use of the proposed system would be able to imp개ve communication among the related participants and systematically accumulate data that might be used in similar construction projects.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.1
• /
• pp.60-70
• /
• 2021

Recently, with the development of technologies for the fourth industrial revolution, convergence and complex technology are being applied to aircraft, electronic home appliances and mobile devices, and the number of parts used is increasing. Increasing the number of parts and the application of convergence technologies such as HW (hardware) and SW (software) are increasing the No Defect Found (NDF) phenomenon in which the defect is not reproduced or the cause of the defect cannot be identified in the subsequent investigation systems after the discovery of the defect in the product. The NDF phenomenon is a major problem when dealing with complex technical systems, and its consequences may be manifested in decreased safety and dependability and increased life cycle costs. Until now, NDF-related prior studies have been mainly focused on the NDF cost estimation, the cause and impact analysis of NDF in qualitative terms. And there have been no specific methodologies or examples of a working-level perspective to reduce NDF. The purpose of this study is to present a practical methodology for reducing NDF phenomena through data mining methods using quantitative data accumulated in the enterprise. In this study, we performed a cluster analysis using market defects and design-related variables of mobile devices. And then, by analyzing the characteristics of groups with high NDF ratios, we presented improvement directions in terms of design and after service policies. This is significant in solving NDF problems from a practical perspective in the company.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.689-696
• /
• 2021

In this paper, in order to apply Prognostics and Health Management(PHM) to an electronic system or circuit, a circuit capable of detecting and predicting defect characteristics inside the system or circuit is implemented, and the results are described. In the previous study, we demonstrated that the frequency of the amplitude of S-parameter changed as the circuit defect progressed. These characteristics were measured by network analyser. but in this study, even if the same defect detection method is used, a circuit is proposed to check the progress of the defect, the remaining time, and the occurrence of the defect without large measurement devices. The circuit is designed to detect the change in impedance that generates changes of S-parameter, and it is verified through simulation using the measurement results of Bond-wires.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.6
• /
• pp.1973-1979
• /
• 2000

Driving the need for machine vision system is growing consumer demand for quality and defect-free products. Especially it is the most important in tarpaulin's manufacturing process achieves automatically by machine vision instead of by man vision. In this paper pinholes detection is performed by using morphology algorithms. Top hat transform is one of morphology applications. This transform take high performance of defect detection in the case that unexpected changes occur in some non-uniform background. For pinholes defect, automatic visual inspection system has been developed, which was composed by a line-scan camera, illumination, a frame grabber, a motor driver and control units. This system has excellent capacity to defect pinholes to the 0.1 mm by 0.5 mm in size and to work in moving objects by maximum 20 m/min in speed.