• 반도체디스플레이기술학회지
• /
• 제13권3호
• /
• pp.1-6
• /
• 2014

Visual defect inspection for electronics parts manufacturing processes is comprised of 2 steps - automatic visual inspection by machine and inspection by human inspectors. It is necessary that spatial points which were detected by the machine should be adequately clustered for subsequent human inspection. This research deals with the spatial clustering algorithm for the purpose of process productivity improvement. Distribution based clustering is newly developed and experimentally confirmed to show better clustering efficiency than existing algorithm - area based clustering.

• 한국멀티미디어학회논문지
• /
• 제15권9호
• /
• pp.1149-1155
• /
• 2012

본 논문에서는 반도체 산업 영상에서 반도체의 결함 원인 진단 기법을 제안한다. 제안 기법은 먼저 결함 영상에 대한 특징 데이터베이스를 구축한다. 다음으로 결함 영상과 입력 영상을 블록 단위로 영역 분할을 수행한 후 컬러 히스토그램을 계산하여 블록들 사이의 히스토그램 카이 제곱 거리를 이용한 블록 유사성을 측정한다. 다음으로 각 영상에서 탐색된 블록들에 대하여 클러스터링을 수행하여 영역을 연결된 객체 단위로 군집한다. 마지막으로 각 클러스터들의 특징을 추출하여 클러스터 간 유사성 측정으로 가장 유사성이 높은 결함 영상을 특징 DB에서 탐색하여 결함 원인 정보와 함께 제시한다. 검색 결과 유사도 상위 n개의 영상 중에서 입력 영상과 동일한 범주의 결함을 갖는 영상이 검색되는 비율을 구하여 제안 기법의 정확성을 검증하였다. n = 1, 2, 3에 대해서 결함 범주에 상관없이 검색 정확도는 모두 100%로 제안 기법은 실제 산업 응용이 가능한 정확한 검색 결과를 보여주었다.

• Journal of Information Processing Systems
• /
• 제13권6호
• /
• pp.1467-1486
• /
• 2017

Fuzzy co-clustering is sensitive to noise data. To overcome this noise sensitivity defect, possibilistic clustering relaxes the constraints in FCM-type fuzzy (co-)clustering. In this paper, we introduce a new possibilistic fuzzy co-clustering algorithm based on information bottleneck (ibPFCC). This algorithm combines fuzzy co-clustering and possibilistic clustering, and formulates an objective function which includes a distance function that employs information bottleneck theory to measure the distance between feature data point and feature cluster centroid. Many experiments were conducted on three datasets and one artificial dataset. Experimental results show that ibPFCC is better than such prominent fuzzy (co-)clustering algorithms as FCM, FCCM, RFCC and FCCI, in terms of accuracy and robustness.

• 한국컴퓨터정보학회논문지
• /
• 제23권9호
• /
• pp.81-86
• /
• 2018

Software defect severity is very important in projects with limited historical data or new projects. But general software defect prediction is very difficult to collect the label information of the training set and cross-project defect prediction must have a lot of data. In this paper, an unclassified data set with defect severity is clustered according to the distribution ratio. And defect severity-based prediction model is proposed by way of labeling. Proposed model is applied CLAMI in JM1, PC4 with the least ambiguity of defect severity-based NASA dataset. And it is evaluated the value of ACC compared to original data. In this study experiment result, proposed model is improved JM1 0.15 (15%), PC4 0.12(12%) than existing defect severity-based prediction models.

• Nuclear Engineering and Technology
• /
• 제50권6호
• /
• pp.907-914
• /
• 2018

In this article, we conducted molecular dynamics simulations to investigate the effect of applied strain and temperature on irradiation-induced damage in alpha-zirconium. Cascade simulations were performed with primary knock-on atom energies ranging between 1 and 20 KeV, hydrostatic and uniaxial strain values ranging from -2% (compression) to 2% (tensile), and temperatures ranging from 100 to 1000 K. Results demonstrated that the number of defects increased when the displacement cascade proceeded under tensile uniaxial hydrostatic strain. In contrast, compressive strain states tended to decrease the defect production rate as compared with the reference no-strain condition. The proportions of vacancy and interstitial clustering increased by approximately 45% and 55% and 25% and 32% for 2% hydrostatic and uniaxial strain systems, respectively, as compared with the unstrained system, whereas both strain fields resulted in a 15-30% decrease in vacancy and interstitial clustering under compressive conditions. Tensile strains, specifically hydrostatic strain, tended to produce larger sized vacancy and interstitial clusters, whereas compressive strain systems did not significantly affect the size of defect clusters as compared with the reference no-strain condition. The influence of the strain system on radiation damage became more significant at lower temperatures because of less annealing than in higher temperature systems.

• Nuclear Engineering and Technology
• /
• 제55권6호
• /
• pp.2298-2304
• /
• 2023

Titanium alloys are expected to become one of the candidate materials for nuclear-powered spacecraft due to their excellent overall performance. Nevertheless, atomistic mechanisms of the defect accumulation and evolution of the materials due to long-term exposure to irradiation remain scarcely understood by far. Here we investigate the heavy irradiation damage in a-titanium with a dose as high as 4.0 canonical displacements per atom (cDPA) using atomistic simulations of Frenkel pair accumulation. Results show that the content of surviving defects increases sharply before 0.04 cDPA and then decreases slowly to stabilize, exhibiting a strong correlation with the system energy. Under the current simulation conditions, the defect clustering fraction may be not directly dependent on the irradiation dose. Compared to vacancies, interstitials are more likely to form clusters, which may further cause the formation of 1/3<1210> interstitial-type dislocation loops extended along the (1010) plane. This study provides an important insight into the understanding of the irradiation damage behaviors for titanium.

• 문화기술의 융합
• /
• 제9권2호
• /
• pp.561-568
• /
• 2023

콘크리트 구조물은 대내외적 환경에 의해 다양한 결함이 발생한다. 결함이 있는 경우 콘크리트의 구조적 안전성에 문제가 있어 이를 효율적으로 파악하여 유지관리하는 것이 중요하다. 하지만, 최근 딥러닝 연구는 콘크리트의 균열에 초점이 맞추어져 있어, 박락과 오염 등에 대한 연구는 부족하다. 본 연구에서는 라벨링이 어려운 박락과 오염에 초점을 맞추어 언라벨 방법, 필터링 방법, 전이학습과 k-means cluster의 융합을 통한 4개의 모델을 개발하고 성능을 평가하였다. 분석결과, 융합모델이 결함을 가장 세밀하게 구분하였으며, 직접 라벨링을 하는 것보다 효율성을 증가시킬 수 있었다. 본 연구 결과가 향후 라벨링이 어려운 다양한 결함 유형에 대한 딥러닝 모델 개발에 기여할 수 있기를 기대한다.

• 산업경영시스템학회지
• /
• 제32권3호
• /
• pp.99-109
• /
• 2009

Negative binomial yield model for semiconductor manufacturing consists of two parameters which are the average number of defects per die and the clustering parameter. Estimating the clustering parameter is quite complex because the parameter has not clear closed form. In this paper, a Bayesian approach using Markov Chain Monte Carlo is proposed to estimate the clustering parameter. To find an appropriate estimation method for the clustering parameter, two typical estimators, the method of moments estimator and the maximum likelihood estimator, and the proposed Bayesian estimator are compared with respect to the mean absolute deviation between the real yield and the estimated yield. Experimental results show that both the proposed Bayesian estimator and the maximum likelihood estimator have excellent performance and the choice of method depends on the purpose of use.

• 정보과학회 컴퓨팅의 실제 논문지
• /
• 제22권12호
• /
• pp.681-686
• /
• 2016

소프트웨어 결함 예측은 프로젝트의 효율적인 관리와 성공에 있어 중요한 요소이다. 이 결함은 심각도에 따라 프로젝트에 영향을 미치는 정도가 다르다. 그러나 기존 연구는 결함 유무만 관심을 두고 심각도를 고려하지 않는다. 본 논문에서는 소프트웨어 관리 효율과 품질 향상을 위해 FCM을 적용한 결함 심각도 기반 앙상블 모델을 제안한다. 제안된 모델은 FCM으로 NASA PC4의 결함심각도를 재분류한다. 그리고 RF(Random Forest)로 심각도에 영향을 주는 입력 column을 선별하여 데이터 핵심 결함 요인을 추출한다. 또한 10-fold 교차검증으로 파라미터를 변경해 모델 성능을 평가한다. 실험 결과는 다음과 같다. 첫째, 결함심각도가 58,40,80에서 30,20,128로 재분류되었다. 둘째, 심각도에 영향을 주는 중요한 입력 column은 정확도와 노드 불순도 측면에서 BRANCH_COUNT였다. 셋째, 성능평가는 트리수가 작고 고려할 변수가 많을수록 좋은 성능을 보였다.

• 대한산업공학회지
• /
• 제40권3호
• /
• pp.257-266
• /
• 2014

The semiconductor manufacturing industry is managed by a number of parameters from the FAB which is the initial step of production to package test which is the final step of production. Various methods for prediction for the quality and yield are required to reduce the production costs caused by a complicated manufacturing process. In order to increase the accuracy of quality prediction, we have to extract the significant features from the large amount of data. In this study, we propose the method for extracting feature from the cell level data of probe test process using OPTICS which is one of the density-based clustering to improve the prediction accuracy of the quality of the assembled chips that will be placed in a package test. Two features extracted by using OPTICS are used as input variables of quality prediction model because of having position information of the cell defect. The package test progress for chips classified to the correct quality grade by performing the improved prediction method is expected to bring the effect of reducing production costs.