• 대한임베디드공학회논문지
• /
• 제17권3호
• /
• pp.129-138
• /
• 2022

Collecting and labeling sufficient training data, which is essential to deep learning-based visual inspection, is difficult for manufacturers to perform because it is very expensive. This paper presents a steel plate surface defect detection system with industrial-grade detection performance by training a small amount of steel plate surface images consisting of labeled and non-labeled data. To overcome the problem of lack of training data, we propose two data augmentation techniques: program-based augmentation, which generates defect images in a geometric way, and generative model-based augmentation, which learns the distribution of labeled data. We also propose a 4-step semi-supervised learning using pseudo labels and consistency training with fixed-size augmentation in order to utilize unlabeled data for training. The proposed technique obtained about 99% defect detection performance for four defect types by using 100 real images including labeled and unlabeled data.

• 산업경영시스템학회지
• /
• 제45권4호
• /
• pp.157-166
• /
• 2022

In the case of a die-casting process, defects that are difficult to confirm by visual inspection, such as shrinkage bubbles, may occur due to an error in maintaining a vacuum state. Since these casting defects are discovered during post-processing operations such as heat treatment or finishing work, they cannot be taken in advance at the casting time, which can cause a large number of defects. In this study, we propose an approach that can predict the occurrence of casting defects by defect type using machine learning technology based on casting parameter data collected from equipment in the die casting process in real time. Die-casting parameter data can basically be collected through the casting equipment controller. In order to perform classification analysis for predicting defects by defect type, labeling of casting parameters must be performed. In this study, first, the defective data set is separated by performing the primary clustering based on the total defect rate obtained during the post-processing. Second, the secondary cluster analysis is performed using the defect rate by type for the separated defect data set, and the labeling task is performed by defect type using the cluster analysis result. Finally, a classification learning model is created by collecting the entire labeled data set, and a real-time monitoring system for defect prediction using LabView and Python was implemented. When a defect is predicted, notification is performed so that the operator can cope with it, such as displaying on the monitoring screen and alarm notification.

• Journal of Periodontal and Implant Science
• /
• 제44권4호
• /
• pp.178-183
• /
• 2014

Purpose: The aim of this study was to assess the possible paracrine effect of bone morphogeneticprotein-2 (BMP-2) at the experimental site on the adjacent control site for validating a rabbit calvarial defect model as a means of verifying the effect of BMP-2. Methods: Sixteen rabbits were divided into two groups (n=8 in each) according to whether or not BMP-2 would be used. Two circular defects (8 mm in diameter) were created side by side, 2 mm apart, in the calvarium of all of the rabbits. In each animal, one of the defects was grafted with either BMP-2-loaded carrier or carrier material alone. The control defects adjacent to these grafted defects, designated CB (the nongrafted defect adjacent BMP-2-loaded carrier-grafted defect) and CC (the nongrafted defect adjacent to carrier only-grafted defect), respectively, were the focus of this study, and were filled only with a blood clot in all of the animals. Histologic observation and histomorphometric analysis were performed at 2 and 8 weeks (n=4 animals per point in time) after surgery. Results: There was no noteworthy difference in the healing pattern, and no statistically significant differences in histomorphometric parameters such as the defect closure, new bone area, or total augmented area between the CC and CB groups. Conclusions: The results of this study suggest that rabbit calvarial defects separated by a distance of 2 mm are suitable for evaluating the effects of BMP-2 and the control defect can be regarded not to be affected by BMP-2 applied defect.

• Steel and Composite Structures
• /
• 제31권3호
• /
• pp.233-242
• /
• 2019

Early detection of damage bonding state such as insufficient bonding strength and interface partial contact defect for the explosive clad pipe is crucial in order to avoid sudden failure and even catastrophic accidents. A generalized and efficient model of the explosive clad pipe can reveal the relationship between bonding state and vibration characteristics, and provide foundations and priory knowledge for bonding state detection by signal processing technique. In this paper, the slender explosive clad pipe is regarded as two parallel elastic beams continuously joined by an elastic layer, and the elastic layer is capable to describe the non-uniform bonding state. By taking the characteristic beam modal functions as the admissible functions, the Rayleigh-Ritz method is employed to derive the dynamic model which enables one to consider inhomogeneous system and any boundary conditions. Then, the proposed model is validated by both numerical results and experiment. Parametric studies are carried out to investigate the effects of bonding strength and the length of partial contact defect on the natural frequency and forced response of the explosive clad pipe. A potential method for identifying the bonding quality of the explosive clad pipe is also discussed in this paper.

• 전기전자학회논문지
• /
• 제26권2호
• /
• pp.319-328
• /
• 2022

베어링은 기계가 작동할때 중요한 역할을 한다. 때문에, 베어링에 결함이 발생하면 기계전체의 치명적인 결함을 발생시킨다. 그러므로 베어링 결함은 조기에 발견되어야한다. 본 논문에서는 연속 웨이블릿 변환과 Switchable 정규화를 기반으로 한 합성곱 신경망(SN-CNN)을 이용한 방법을 베어링 결함 감지 모델에 대해 설명한다. 모델의 정확도는 Case Western Reserve University(CWRU) 베어링 데이터 집합을 사용하여 측정되었다. 또한 배치 정규화(BN, Batch Normalization)[1] 방법과 스펙트로그램 이미지가 모델 성능의 비교를 위해 사용되었다.

• 한국해양공학회지
• /
• 제13권4호통권35호
• /
• pp.45-54
• /
• 1999

Non destructive test is applied to revise mechanical strength and assume material strength or defect of material, equipment and structure, instead of fracture test. Especially, ultrasonic test has the characteristics such as an excellent permeability high-sensitiveness to fine defect and an almost exact measurement for position, size and direction of inner defect which differ from other non destructive tests. In this study, the program is developed to evaluate optimal testing condition, to distinguish obstacle echo and defect position. This program on the basic of Ray-Tracing model shows generation and processing of ultrasonic pulse. The simulation is compared with testing in the 3 cases of an oblique angle transducer like $45^{\circ},\;60^{\circ}\;and\;70^{\circ}$. The test result for all conditions is well compared with simulation result when relative not is within $0.1{\sim}7.2%$. And the course of several echos is simply assumed through simulation.

• International Journal of Computer Science & Network Security
• /
• 제22권10호
• /
• pp.11-16
• /
• 2022

Applying predictive analytics to predict software defects has improved the overall quality and decreased maintenance costs. Many supervised and unsupervised learning algorithms have been used for defect prediction on publicly available datasets. Most of these datasets suffer from an imbalance in the output classes. We study the impact of class imbalance in the defect datasets on the efficiency of the defect prediction model and propose a CPP method for handling imbalances in the dataset. The performance of the methods is evaluated using measures like Matthew's Correlation Coefficient (MCC), Recall, and Accuracy measures. The proposed sampling technique shows significant improvement in the efficiency of the classifier in predicting defects.

• 반도체디스플레이기술학회지
• /
• 제23권2호
• /
• pp.135-142
• /
• 2024

Cu-Cu bonding, one of the key technologies in advanced packaging, enhances semiconductor chip performance, miniaturization, and energy efficiency by facilitating rapid data transfer and low power consumption. However, the quality of the interface bonding can significantly impact overall bond quality, necessitating strategies to quickly detect and classify in-process defects. This study presents a methodology for detecting defects in wafer junction areas from Scanning Acoustic Microscopy images using a ResNet-50 based deep learning model. Additionally, the use of the defect map is proposed to rapidly inspect and categorize defects occurring during the Cu-Cu bonding process, thereby improving yield and productivity in semiconductor manufacturing.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 하계학술대회 논문집 C
• /
• pp.2006-2008
• /
• 2004

In this paper the imitated electric defect was simulated by insulation particle in the model GIS. This paper studied the distribution of frequency spectrum of the radiated electromagnetic waves using antenna (30-2,000[MHz]) and spectrum analyzer. From results of this study, a new method was introduced for measurement and analysis of the radiated electromagnetic waves in accordance with discharge progress of each defect in the model GIS. It was confirmed that detecting partial discharge and estimating discharge progress can be possible in the model GIS.

• 한국연소학회지
• /
• 제4권2호
• /
• pp.51-62
• /
• 1999

NOx formation in turbulent flames is strongly coupled with temperature, superequilibrium concentration of O radical, and residence time. This implies that in order to accurately predict NO level, it is necessary to develop sophisticated models able to account for the complex turbulent combustion processes including turbulence/chemistry interaction and radiative heat transfer. The present study numerically investigates the turbulent nonpremixed hydrogen jet flames using the laminar flamelet model. Flamelet library is constructed by solving the modified Peters equations and the turbulent combustion model is extended to nonadiabatic flame by introducing the enthalpy defect. The effects of turbulent fluctuation are taken into account by the presumed joint PDFs for mixture fraction, scalar dissipation rate, and enthalpy defect. The predictive capability of the present model has been validated against the detailed experimental data. Effects of nonequilibrium chemistry and radiative heat loss on the thermal NO formation are discussed in detail.