• 대한전자공학회논문지SP
• /
• 제48권3호
• /
• pp.19-29
• /
• 2011

지금까지 많은 얼굴 인식 방법들이 제안되었으나, 대부분의 방법들은 특징추출 과정 없이 입력 영상을 1차원 형태의 벡터로 변형한 것을 1차원 특징 벡터로 사용하거나 또는 입력 영상 자체를 특징 매트릭스로 사용하였다. 이와같이 영상 자체를 특징으로 사용하면 조명변화가 심한 데이터베이스에서는 성능이 좋지 않는 것으로 알려져 있다. 본 논문에서는 조명변화에 효과적인 그래디언트와 상관관계의 국부통계를 이용하여 얼굴을 인식하는 방법을 제안하였다. BDIP(block difference of inverse probabilities)는 그래디언트의 국부 통계이다. 그리고 BVLC(block variation of local correlation coefficients)의 두 타입은 상관관계의 국부 통계이다. 입력영상이 얼굴인식 시스템에 들어 오면 먼저 BDIP, BVLC1, BVLC2의 특징 영상을 추출하고 융합한 후, (2D)2 PCA 변환을 거쳐 특징 매트릭스를 얻어서 훈련특징 매트릭스와의 거리를 구하여 최근린 분류기를 이용하여 얼굴 영상을 인식한다. 네 가지 얼굴 데이터베이스, FERET, Weizmann, Yale B, Yale에 대한 실험결과로부터, 제안한 방법이 실험한 여섯 가지 방법 중에서 조명과 얼굴 표정의 변화에 가장 견실하다는 것을 알 수 있었다.

• 한국정보통신학회논문지
• /
• 제8권1호
• /
• pp.123-136
• /
• 2004

본 논문에서는 $GF({2^n})$상 곱셈의 복잡도와 규칙도를 GF(2)상의 다항식 곱셈을 표현하는 행렬식의 행과 열의 해밍 가중치를 이용하여 정의한다 차분공격에 강한 블록 암호 알고리즘을 만들기 위해서는 치환계층과 확산계층의 $GF({2^n})$상 곱셈의 복잡도와 규칙도가 높아야함을 실험을 통하여 보인다. 실험 결과를 활용하여 우리나라 표준인 128 비트 블록 암호 알고리즘인 SEED의 S 박스와 G 함수를 구성하는 방식을 제안한다. S 박스는 비 선형함수와 아핀변환으로 구성한다. 비 선형함수는 차분공격과 선형공격에 강한 특성을 가지며, '0'과 '1'을 제외하고 입력과 출력이 같은 고정점과 출력이 입력의 1의 보수가 되는 역고정점을 가지지 않는 $GF({2^8})$ 상의 역수로 구성한다. 아핀변환은 입력과 출력간의 상관을 최저로 하면서 고정점과 역고정점이 없도록 구성한다. G 함수는 4개의 S 박스 출력을 $GF({2^8}) 상의 4 {\times} 4$ 행렬식을 사용하여 선형변환한다. 선형변환 행렬식 성분은 높은 복잡도와 규칙도를 가지도록 구성한다 또한 MDS(Maximum Distance Separable) 코드를 생성하고, SAC(Strict Avalanche Criterion)를 만족하고, 고정점과 역고정점 및 출력이 입력의 2의 보수가 되는 약한 입력이 없도록 G 함수를 구성한다. 비선형함수와 아핀변환 및 G 함수의 원시다항식은 각기 다른 것을 사용한다. 본 논문에서 제안한 S 박스와 G 함수는 차분공격과 선형공격에 강하고, 약한 입력이 없으며, 확산 특성이 우수하므로 안전성이 높은 암호 방식의 구성 요소로 활용할 수 있다.

• 한국습지학회지
• /
• 제9권1호
• /
• pp.107-114
• /
• 2007

This study aimed to investigate the hexachlorobenzene (HCB) dechlorinating ability of sediment microbes collected from a natural canal receiving secondary effluents from an industrial estate and nearby factories. Nine sites along the stream and one in the estuary in the Gulf of Thailand into which the canal spills were specified and sampling for sediment and water. Preliminary analysis of the sediments showed that the first four sites nearest to the discharging location were contaminated by HCB within the range of 0.18 to 1.25 ppm. Apart from that, 1,3,5-trichlorobenzene which has never been commercially produced or used in any manufacturing processes except for the transformation from higher chlorinated benzene was also identified in the range of 0.16 to 0.24 ppm. This suggested a possibility of sporadically HCB contamination in this stream. Of more important, people in the community along this canal earn their living by coastal fishery; hence, posing a risk of spreading HCB and its less chlorinated congeners via food chain from caught marine creatures to human. As a result, there is an urgent need to understand the behavior of HCB dechlorination in this stream sediment which can lead to a clean-up action in the future. Serum bottles with sediment slurries (sediment to water ratio of 1:1 (v/v) and filtered to remove particles larger than 0.7 mm) from each site were inoculated with 2 mg/l of HCB, kept anaerobically in the dark at room temperature without any nourishment, and analyzed for HCB and its less-chlorinated congeners every 6 days. Total chemical oxygen demand, suspended solids, and volatile suspended solids were in the range of 21,492-73,584, 158,100-518,100 and 6,000-32,700 mg/l, respectively. It was found that all sediment slurries began to dechlorinate HCB in 12 to 30 days and the HCB was completely removed within 42 to 60 days or so. On the other hand, there was no HCB dechlorination occurred in the controlled set which was sterilized by autoclaving prior to the addition of HCB. This implies that the HCB transformation was solely due to microorganisms' activities. HCB was dechlorinated principally via pentachlolobenzene to 1,2,3,5-tetrachlorobenzene and terminated at 1,3,5-trichlorobenzene which is the major pathway as reported by many researchers. Dichlorobenzene has not been detected in any samples within the dechlorination period of 60 days. The results indicate that the microbial matrix in the sediment of this stream has an outstanding capability to dechlorinate HCB. Existing substrates and nutrients which mainly sorbed onto the solid phase and the typical temperature in Thailand were sufficient and suitable to promote the activities of these HCB-dechlorinating microbes.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2008년도 학술발표회 논문집
• /
• pp.694-698
• /
• 2008

In hydrology, it is a difficult task to deal with multivariate time series such as modeling streamflows of an entire complex river system. Normal distribution based model such as MARMA (Multivariate Autorgressive Moving average) has been a major approach for modeling the multivariate time series. There are some limitations for the normal based models. One of them might be the unfavorable data-transformation forcing that the data follow the normal distribution. Furthermore, the high dimension multivariate model requires the very large parameter matrix. As an alternative, one might be decomposing the multivariate data into independent components and modeling it individually. In 1985, Lins used Principal Component Analysis (PCA). The five scores, the decomposed data from the original data, were taken and were formulated individually. The one of the five scores were modeled with AR-2 while the others are modeled with AR-1 model. From the time series analysis using the scores of the five components, he noted "principal component time series might provide a relatively simple and meaningful alternative to conventional large MARMA models". This study is inspired from the researcher's quote to develop a multivariate simulation model. The multivariate simulation model is suggested here using Principal Component Analysis (PCA) and Independent Component Analysis (ICA). Three modeling step is applied for simulation. (1) PCA is used to decompose the correlated multivariate data into the uncorrelated data while ICA decomposes the data into independent components. Here, the autocorrelation structure of the decomposed data is still dominant, which is inherited from the data of the original domain. (2) Each component is resampled by block bootstrapping or K-nearest neighbor. (3) The resampled components bring back to original domain. From using the suggested approach one might expect that a) the simulated data are different with the historical data, b) no data transformation is required (in case of ICA), c) a complex system can be decomposed into independent component and modeled individually. The model with PCA and ICA are compared with the various statistics such as the basic statistics (mean, standard deviation, skewness, autocorrelation), and reservoir-related statistics, kernel density estimate.

• 대한의용생체공학회:의공학회지
• /
• 제24권3호
• /
• pp.183-192
• /
• 2003

본 연구에서는 구현된 요분석 시스템으로 획득한 데이터를 보정하기 위하여 색 좌표 변환 기법을 제안하였다. 일반적으로 요분석 시스템은 요분석용 스트립의 정색반응을 검출하는 과정에서 여러 가지 비선형적인 특성 즉 광 모듈의 메커니즘, 하드웨어, 그리고 주변 환경에 의해 색 왜곡된 입·출력 특성을 지닌다 따라서 보다 높은 정확도와 재현성을 유지하기 위해 장비 특성화 기법을 도입하여 색 왜곡 현상을 보정하였다. 본 연구에서는 명암 보정, ,3차 스플라인 보간법에 의한 RGB 신호의 특성 곡선 추출, 기준색 고정 선형변환 기법을 사용하여 색 보정 과정을 수행하였다 색 보정을 위해 사용된 표준 장비는 1931년 CIE XYZ 색공간 특성을 지닌 좌표계로 설정하였으며, 동일한 칼라 샘플에 대해 요분석 시스템의 출력값과 표준 장비의 출력값이 일치되도록 하는 보정 행렬을 구하였다. 구현된 요분석 시스템을 색 보정한 후 기준 데이터와 비교한 결과 양호한 색상 정확도를 나타내었다 요분석용 스트립의 10 가지 항목에 대해 구현된 두 대의 요분석 시스템을 사용하여 실험한 결과 장비간 색차는 1.28이었다.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• 제27권2호
• /
• pp.129-134
• /
• 2001

Since NNK is one of the most abundant tobacco-specific alkaloids and a strong carcinogenic nitrosamine, it has been used for evaluating a potential of carcinogenicity in the animal models. The present study has attempted to examine the potential of carcinogenicity of NNK in human epithelial cells, from which the cell type the most of cancers including oral cancer and nasal cavity cancer are originated. The cellular model used for the study is a human keratinocyte cell system immortalized by Ad12-SV40 hybrid virus. The cellular system has successfully been used for the carcinogenicity studies because of its limitless life span, epithelial morphology and nontumorigenicity. When cells were treated with a variety of NNK concentrations, levels of saturation density and soft agar colony formation were increased in a dose-dependent fashion. Colonies of large cell aggregates were above 5 at the higher doses. The results indicate that exposure of human cells with NNK induced loss of contact inhibition and increases of anchorage independence and cellular adhesion, which are typical characteristics of the neoplatically transformed cells. When cells were exposed with 100uM NNK for 2hr, mRNA levels of IL-1 and PAI-2 were increased in a dose-dependent manner, but expression of TGF- 1 was not affected. While expression of growth regulatory factors were altered with a short-term exposure, there was no alteration of these factors in the NNK-transformed cells. However, mRNA levels of fibronectin were increased both in the short-term treatment and in the transformation. The results suggest that altered expression of extracellular matrix such as fibronectin following short-term exposure might be fixed in the genome and these altered properties be continuously transfered throughout the cell division. Western blot analysis showed a translocation of PKC- from cytosolic fraction to the particulate fraction, indicating a possible role of NNK in the signal transduction pathway. The present study provided an evidence that NNK in the smoking may be associated with epithelial origin cancer such as oral and nasal cavity cancers. In addition, this study suggested that altered expression of extracellular matrix and PKC may play an important role in the carcinogenic mechanism of NNK.

• Journal of Welding and Joining
• /
• 제33권5호
• /
• pp.53-60
• /
• 2015

Laser surface Melting Process is getting hardening layer that has enough depth of hardening layer as well as no defects by melting surface of substrate. This study used CW(Continuous Wave) Yb:YAG and STD11. Laser beam speed, power and beam interval are fixed at 70mm/sec, 2.8kW and 800um respectively. Hardness in the weld zone are equal to 400Hv regardless of melting zone, remelting zone overlapped by next beam and HAZ. Similarly, microstructures in all weld zone consist of dendrite structure that arm spacing is $3{\sim}4{\mu}m$, matrix is ${\gamma}$(Austenite) and dendrite boundary consists of ${\gamma}$ and $M_7C_3$ of eutectic phase. This microstructure crystallizes from liquid to ${\gamma}$ of primary crystal and residual liquid forms ${\gamma}$ and $M_7C_3$ of eutectic phase by eutectic reaction at $1266^{\circ}C$. After solidification is complete, primary crystal and eutectic phase remain at room temperature without phase transformation by quenching. On the other hand, microstructures of substrate consist of ferrite, fine $M_{23}C_6$ and coarse $M_7C_3$ that have 210Hv. Microstructures in the HAZ consist of fine $M_{23}C_6$ and coarse $M_7C_3$ like substrate. But, $M_{23}C_6$ increases and matrix was changed from ferrite to bainite that has hardness above 400Hv. Partial Melted Zone is formed between melting zone and HAZ. Partial Melted Zone near the melting zone consists of ${\gamma}$, $M_7C_3$ and martensite and Partial Melted Zone near the HAZ consists of eutectic phase around ${\gamma}$ and $M_7C_3$. Hardness is maximum 557Hv in the partial melted zone.

• 한국분말재료학회지
• /
• 제13권3호
• /
• pp.178-186
• /
• 2006

The effect of use of $H_2O$ as PCA(process control agent) to prevent the carbon contamination during mechanical alloying process and the precipitation behavior in Ni-20Cr-20Fe-5Nb bulk alloy after aging were investigated. NbC and $Cr_2O_3$ were formed during mechanical alloying and consolidation processes in the Ni-20Cr-20Fe-5Nb alloy in which methanol($CH_3OH$) was added as PCA. Formation of NbC in this alloy decreased the amount of Nb dissolved in the Ni matrix. The use of $H_2O$ as PCA in Ni-20Cr-20Fe-5Nb alloy prevented the formation of NbC and increased the hardness. The increase of hardness in this alloy was attributed to the increased amount of Nb dissolved in the Ni matrix. After aging treatment for 20 hours at $600^{\circ}C\;and\;720^{\circ}C$ of Ni-20Cr-20Fe-5Nb bulk alloy in which $H_2O$ added as PCA, ${\gamma}"$$(Ni_3Nb,\;tetragonal)\;and\;{\delta}\;(Ni_3Nb,\;orthorhombic)$ precipitates were formed, respectively. The precipitation temperatures of ${\gamma}"$ and ${\delta}$ in this bulk alloy were lower than those in commercial IN 718 alloy. It seemed that the lower precipitation temperatures for ${\gamma}"$ and ${\delta}$ in this bulk alloy than in commercial IN 718 alloy were due to severe plastic deformation during mechanical alloying.

• 대한전자공학회논문지SP
• /
• 제39권4호
• /
• pp.429-441
• /
• 2002

본 논문은 실시간으로 3차원 공간상에서의 움직임 정보를 추출할 수 있는 입력 장치를 제안한다. 제안하는 3차원 입력 장치는 스테레오 카메라의 기하학적 구조와 색상, 움직임, 형태상의 특성을 이용하여 복잡한 환경에서 사전 카메라 캘리브레이션 없이 3차원 움직임 정보를 추출할 수 있다. 움직임 추출을 위해서 perspepctive projection 행렬과 perspective distortion 행렬을 이용한 스테레오 카메라의 기하학적 특성을 이용하며, 효과적인 좌우 영상의 특징점 추적 및 추출을 위해 색상 변환(Color transform)과 UPC(Unmatched Pixel Count) 및 이산 칼만 필터(Discrete Kalman Filter)의 효과적인 결합으로 이루어진 MAWUPC(Motion Adaptive Weighted Pixel Count)과 PCA(Principal Component Analysis)로 구성된 알고리즘을 제안한다. 추출된 3차원 공간상에서의 움직임은 가상환경에서의 가상 물체를 제어하거나 사용자 시점의 이동을 나타내는 인터페이스로 사용한다. 스테레오 비전을 이용한 입력 장치는 선으로 연결되지 않기 때문에 사용자가 가상환경에서 작업하기가 편리하며 몰입감을 높일 수 있는 등 보다 효율적인 상호작용을 가능하게 해준다.

• 대한토목학회논문집
• /
• 제14권6호
• /
• pp.1299-1308
• /
• 1994

본 연구는 사장교의 차량하중에 의한 동적거동을 파악하고자 수치해석상 비교적 간단한 일본의 풍리(豊里)(Toyosato)대교(大橋)의 자료를 근거로하여 수치해석 대상모델을 구조형식별로, 여러가지 설계변수-즉, (1) 경간비, (2) 중앙경간장과 주탑높이와의 비, (3) 거어더의 강성, (4) 주탑의 강성, (5) 케이블의 강성-을 변화시켜 수치해석을 수행하여 동적거동을 파악하고, 그 결과를 가지고 설계변수의 영향 및 충격계수의 변화에 대하여 비교 분석하였다. 이때 변위 및 단면력의 영향선을 구하기 위한 해석은 전달행렬법을 이용하였으며, 동적해석에 있어서는 평면구조계의 집중질량계로 모델을 가정하여 차량과 교량의 운동방정식을 유도한 후 모드중첩법을 사용하여 각 질점에 대한 변위 및 단면력의 동적시간이력을 구하였다.