• 대한임베디드공학회논문지
• /
• 제10권5호
• /
• pp.265-272
• /
• 2015

This paper presents an improved feature extraction method to achieve a good performance in the detection of splicing forged images. Strong edges caused by the image splicing destroy the statistical dependencies between parent and child subbands in the wavelet domain. We analyze the co-occurrence probability matrix of parent and child subbands in the wavelet domain, and calculate the statistical moments from two-dimensional characteristic function of the co-occurrence matrix. The extracted features are used as the input of SVM classifier. Experimental results show that the proposed method obtains a good performance with a small number of features compared to the existing methods.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 정보 및 제어부문
• /
• pp.9-11
• /
• 2005

We propose a new content-based image retrieval using a block color co-occurrence matrix (BCCM) and pattern correlogram. In the proposed method, the color feature vectors are extracted by using BCCM that represents the probability of the co-occurrence of two mean colors within blocks. Also the pattern feature vectors are extracted by using pattern correlogram which is combined with spatial correlation of pattern. In the proposed pattern correlogram method. after block-divided image is classified into 48 patterns with respect to the change of the RGB color of the image, joint probability between the same pattern from the surrounding blocks existing at the fixed distance and the center pattern is calculated. Experimental results show that the proposed method can outperform the conventional methods as regards the precision and the size of the feature vector dimension.

• 한국정보통신학회논문지
• /
• 제20권4호
• /
• pp.833-839
• /
• 2016

본 논문에서는 효율적인 접합 영상 검출을 위한 마르코프 천이 및 동시발생 확률에 대한 선택적 특징 추출 방법을 제안한다. 제안하는 방법에서는 이산 코사인 변환 영역에서 블록간 계수의 차이를 이용하여 특징들을 구성하고, 특징들의 각 위치에서 원 영상과 접합영상의 특징 분포의 상이성을 확인하기 위해 Kullback-Leibler 수렴값을 구한다. 이를 바탕으로, 마르코프 확률 특징과 동시발생 확률 특징 가운데 해당 위치에서 가장 큰 차이값을 갖는 특징을 선택하여 최종 특징으로 선택하고, SVM 분류기를 이용하여 학습 및 테스트한 후 그 유효성을 판별한다. 실험 결과를 바탕으로 제안하는 방법이 기존의 방법보다 제한된 특징수로 높은 영상접합 조작 결과를 보임을 확인하였다.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제15권6호
• /
• pp.405-413
• /
• 2009

태깅 시스템은 인터넷 사용자로 하여금 태그라고 불리는 메타데이터를 글, 사진, 동영상 등에 부여하도록 하여서 컨텐츠의 검색 및 브라우징을 편리하게 하는 시스템이다. 컨텐츠의 브라우징을 위해서 태그 클라우드라는 시각적 인터페이스가 널리 쓰이고 있다. 태그 클라우드는 가장 빈도수가 높은 태그들을 알파벳 순으로 보여주고 폰트의 크기로 그 태그들의 빈도수를 반영한다. 하지만 기존의 태그 선택 방법은 몇 가지 단점들이 알려져 있다. 그래서 이 논문은 참신한 컨텐츠들을 찾을 수 있도록 Freshness라는 태그 클라우드를 위한 새로운 태그 선택 방법을 정의하였다. Freshness는 태그 동시 발생 확률 분포(tag co-occurrence probability distribution)가 동적으로 변화하는 것을 Kullback-Leibler divergence로 평균한 값이다. Allblog, Eolin, Technorati 등 세 개의 웹사이트로부터 실제 태그 데이터를 수집하여 우리의 태그 클라우드를 생성하는 시스템, 'Fresh Tag Cloud'를 구축하였다. 이 태그 클라우드를 Allblog에서 수집한 데이터에서 전통적인 태그 클라우드와 비교했을 때 중복평균이 87.5% 감소하여서 성능이 더 향상된 것을 확인할 수 있다.

• Korean Journal of Acupuncture
• /
• 제33권1호
• /
• pp.18-32
• /
• 2016

Objectives : This study aims to analyze the co-occurrence of pathological symptoms and corresponding acupoints as documented by the comprehensive acupuncture and moxibustion records in the classical texts of Far East traditional medicine as an aid to a more efficient understanding of the tacit treatment principles of ancient physicians. Methods : The Classic of Nourishing Life with Acupuncture and Moxibustion(Zhenjiu Zisheng Jing; hereinafter ZZJ) was selected as the primary reference book for the analysis. The pathology-acupoint co-occurrence analysis was performed by applying 4 values of vector space measures(weighted Euclidean distance, Euclidean distance, $Cram\acute{e}r^{\prime}s$ V and Canberra distance), which measure the distance between the observed and expected co-occurrence counts, and 3 values of probabilistic measures(association strength, Fisher's exact test and Jaccard similarity), which measure the probability of observed co-occurrences. Results : The treatment records contained in ZZJ were preprocessed, which yielded 4162 pathology-acupoint sets. Co-occurrence was performed applying 7 different analysis variables, followed by a prediction simulation. The prediction simulation results revealed the Weighted Euclidean distance had the highest prediction rate with 24.32%, followed by Canberra distance(23.14%) and association strength(21.29%). Conclusions : The weighted Euclidean distance among the vector space measures and the association strength among the probabilistic measures were verified to be the most efficient analysis methods in analyzing the correlation between acupoints and pathologies found in the classical medical texts.

• 한국멀티미디어학회논문지
• /
• 제20권1호
• /
• pp.43-50
• /
• 2017

According to living environment, the number of deaths due to thyroid diseases increased. In this paper, we proposed an algorithm for recognizing a thyroid detection using texture analysis based on shape, gray level co-occurrence matrix and gray level run length matrix. First of all, we segmented the region of interest (ROI) using active contour model algorithm. Then, we applied a total of 18 features (5 first order descriptors, 10 Gray level co-occurrence matrix features(GLCM), 2 Gray level run length matrix features and shape feature) to each thyroid region of interest. The extracted features are used as statistical analysis. Our results show that first order statistics (Skewness, Entropy, Energy, Smoothness), GLCM (Correlation, Contrast, Energy, Entropy, Difference variance, Difference Entropy, Homogeneity, Maximum Probability, Sum average, Sum entropy), GLRLM features and shape feature helped to distinguish thyroid benign and malignant. This algorithm will be helpful to diagnose of thyroid nodule on ultrasound images.

• Journal of Advanced Research in Ocean Engineering
• /
• 제1권4호
• /
• pp.239-251
• /
• 2015

In this study, a spectral fatigue analysis was performed for the topside structure of an offshore floating vessel. The topside structure was idealized using beam elements in the SACS program. The fatigue analysis was carried out considering the wave and wind loads separately. For the wave-induced fatigue damage calculation, motion RAOs calculated from a direct wave load analysis and regular waves with different periods and unit wave heights were utilized. Then, the member end force transfer functions were generated covering all the loading conditions. Stress response transfer functions at each joint were produced using the specified SCFs and member end force transfer functions. fatigue damages were calculated using the obtained stress ranges, S-N curve, wave spectrum, heading probability of each loading condition, and their corresponding occurrences in the wave scatter diagrams. For the wind induced fatigue damage calculation, a dynamic wind spectral fatigue analysis was performed. First, a dynamic natural frequency analysis was performed to generate the structural dynamic characteristics, including the eigenvalues (natural frequencies), eigenvectors (mode shapes), and mass matrix. To adequately represent the dynamic characteristic of the structure, the number of modes was appropriately determined in the lateral direction. Second, a wind spectral fatigue analysis was performed using the mode shapes and mass data obtained from the previous results. In this analysis, the Weibull distribution of the wind speed occurrence, occurrence probability in each direction, damping coefficient, S-N curves, and SCF of each joint were defined and used. In particular, the wind fatigue damages were calculated under the assumption that the stress ranges followed a Rayleigh distribution. The total fatigue damages were calculated from the combination with wind and wave fatigue damages according to the DNV rule.

• Wind and Structures
• /
• 제19권5호
• /
• pp.467-480
• /
• 2014

The estimation of wind speed values used in codes and standards is an integral part of the wind load evaluation process. In a number of codes and standards, wind speeds outside of tropical cyclone prone regions are estimated using a single probability distribution developed from observed wind speed data, with no distinction made between the types of causal wind hazard (e.g., thunderstorm). Non-tropical cyclone wind hazards (i.e., thunderstorm, non-thunderstorm) have been shown to possess different probability distributions and estimation of non-tropical cyclone wind speeds based on a single probability distribution has been shown to underestimate wind speeds. Current treatment of non-tropical cyclone wind hazards in worldwide codes and standards is touched upon in this work. Meteorological data is available at a considerable number of United States (U.S.) stations that have information on wind speed as well as the type of causal wind hazard. In this paper, probability distributions are fit to distinct storm types (i.e., thunderstorm and non-thunderstorm) and the results of these distributions are compared to fitting a single probability distribution to all data regardless of storm type (i.e., co-mingled). Distributions fitted to data separated by storm type and co-mingled data will also be compared to a derived (i.e., "mixed") probability distribution considering multiple storm types independently. This paper will analyze two extreme value distributions (e.g., Gumbel, generalized Pareto). It is shown that mixed probability distribution, on average, is a more conservative measure for extreme wind speed estimation. Using a mixed distribution is especially conservative in situations where a given wind speed value for either storm type has a similar probability of occurrence, and/or when a less frequent storm type produces the highest overall wind speeds. U.S. areas prone to multiple non-tropical cyclone wind hazards are identified.

• 지능정보연구
• /
• 제8권2호
• /
• pp.1-13
• /
• 2002

본 논문에서는 연속적으로 입력되는 칼라영상에서 물체의 이동에 의하여 형성된 동작영역을 확인하고, 영상의 시컨스(sequence)를 대상으로 움직이는 물체의 형태인 보행자 혹은 자동차들의 이동방향을 추적하는 시스템을 제안하였다. 카메라가 고정되어 있고 물체가 이동하는 상황에서 카메라시계에 진입하는 물체를 포착하여, 포착된 물체의 영역을 차 영상 분석을 통해 이진화하여 추출하고, 추출된 영역을 co-occurrence matrix의 RGB full 칼라의 특징 벡터를 추출하는 것을 제시하였다 추출되어지는 칼라 특징벡터를 분석하여 인접 프레임간의 이동물체 영역끼리의 대응관계를 조사함으로서, 이동물체를 추적한다. 군집화(clustering) 단계에서는 이전 단계에서 추출한 특징 벡터들 가운데 에너지, 엔트로피만을 가지고 인접 프레임간의 군집화를 조사하기 위하여 이동물체 영역들 간의 퍼지동적물체 정합 알고리즘을 적용시켰다. 인접 프레임간의 움직임 영역의 물체들에 대하여 멤버 쉽 함수를 근거로 중심 값을 계산하면, 동일 물체일 경우 중심 값 부근에서 군집이 형성되며, 이를 바탕으로 이동물체를 추출할 수 있는 방안을 제안하였다.

• 대한전자공학회논문지SD
• /
• 제43권8호
• /
• pp.35-43
• /
• 2006

이 논문에서는 공정 변이의 고려를 위한 통계적 시간 분석(statistical timing analysis)에서 전력감소를 고려한 회로의 최적화를 위해 글리치 및 지연시간의 확률적 모델 및 연산을 이용하여 각 경로 및 경로상의 게이트의 민감도(sensitivity)를 계산하고 이를 이용한 사이징(sizing)을 통해 회로의 지연시간의 증가 없이 글리치를 감소하는 방법을 제시한다. 제안된 알고리즘은 통계적 시간 분석에 근거한 회로의 전후방 탐색을 이용하여 공정 변수를 고려한 확률적 글리치 발생률을 예측한다. 또한 글리치 발생률을 고려한 게이트의 선택 및 사이징 가능한 지연시간의 최적화된 계산을 통해 효율적인 게이트 사이징 기법과 글리치 감소를 위한 경로균등화 방법을 제시한다. 제안된 알고리즘의 효율성은 $0.16{\mu}m$ 모델 파라미터를 이용하여 ISCAS85 벤치마크 회로에 대한 실험을 통해 검증되었다. 실험 결과를 통해 제안된 알고리즘은 글리치 예측에 있어 8.6%의 정확도의 개선을 보였고, 경로균등화에 의한 최적화에 있어 9.5%의 개선을 보였다.