• 응용통계연구
• /
• 제30권2호
• /
• pp.203-213
• /
• 2017

우리는 이변량 경시적 자료의 조건부 결합 분포를 추정하기 위하여 회귀 모형과 코플라 모형을 연구하였다. 주변 분포의 추정을 위하여 시변 변환 모형을 고려하였고, 이변량 반응변수 각각에 대한 주변 분포를 가우시안 코플라를 이용하여 결합하여 조건부 결합 분포를 추정하였다. 우리가 제안한 모형은 조건부 평균 모형만으로 자료를 설명하기 어려운 경우에 적용될 수 있다. 시변 변환 모형과 가우시안 코플라 모형을 결합한 본 논문의 방법은 반복 측정된 이변량 경시적 자료에 대한 모형화가 용이하며 해석하기 쉬운 장점이 있다. 우리는 본 논문의 방법을 반복 측정된 이변량 콜레스테롤 자료를 분석하는데 적용하여 보았다.

• Wind and Structures
• /
• 제24권1호
• /
• pp.59-78
• /
• 2017

Non-stationarity and non-Gaussian property are two of the most important characteristics of wind. These two features are studied in this study based on wind speed records measured at different heights from a 325 m high meteorological tower during the synoptic wind storms. By using the time-frequency analysis tools, it is found that after removing the low frequency trend of the longitudinal wind, the retained fluctuating wind speeds remain to be asymmetrically non-Gaussian distributed. Results show that such non-Gaussianity is due to the weak-stationarity of the detrended fluctuating wind speed. The low frequency components of the fluctuating wind speeds mainly contribute to the non-zero skewness, while distribution of the high frequency component is found to have high kurtosis values. By further studying the decomposed wind speed, the mechanisms of the non-Gaussian distribution are examined from the phase, turbulence energy point of view.

• 한국통신학회논문지
• /
• 제25권1B호
• /
• pp.99-104
• /
• 2000

본 논문은에서는 저작권 보호를 위한 디지털 워터마크 삽입방법에서 워터마크로 많이 사용하는 정규 가우시안 시퀀스를 카오스 시퀀스로 대체하고 그 성능을 비교하여 분석한다. 카오스 시퀀스는 만들기가 쉽고, 초기 치의 변화에 따라 전혀 다른 시퀀스를 만들 수 있다. 본 논문에서 사용한 카오스 시퀀스는 Chebyshev map의 시퀀스 분포를 갖도록 Logistic map을 수정하였다. 실험방법은 원 영상을 웨이브릿 변환하여 카오스 시퀀스와 가우시안 시퀀스로 워터마킹한 후 여러 가지 영상처리와, 반복적인 실험의 결과로 나타난 유사도의 분포를 측정, 비교하였다. DCT-기반 워터마킹 시스템의 결과와 마찬가지로 카오스 시퀀스는 일반적인 신호처리에 있어서 가우시안 시퀀스 못지 않게 강하다. 또한 연속적인 반복 실험에 의한 유사도 편차가 가우시안의 경우보다 작고, 손실 압축에 있어서는 가우시안 시퀀스 보다 좋은 성능을 보였다.

• 디지털산업정보학회논문지
• /
• 제7권4호
• /
• pp.81-88
• /
• 2011

Mini Double Gears Frame is critical part of PDP and also produces couple hundred thousand every month. In the process of mass production, product inspection is very important process. Double Gear, one of the part of machine, was inspected by human eyes which caused mistakes and slow progress. To achieve the speed and accuracy the system was compensated by vision system which is inspecting automatically. The focus value is measured based on the fact that high contrast images have much high frequency edge information. High frequency term of the image is extracted using the high-pass filter and the sum of the high frequency term is used as the focus value. We used a Gaussian smoothing filter to reduce the noise and then measures the focus value using the modified Laplacian filter called a Sum modified Laplacian Focus values for the various lens positions are calculated and the position with the maximum focus value is decided as the focused position. The focus values calculated in various lens position showed the Gaussian distribution. We proposed a method to estimate the best focus position using the Gaussian curve fitting. Focus values of the uniform interval lens positions are calculated and the values are used to estimate the Gaussian distribution parameters to find the best focus position.

• International Journal of Advanced Culture Technology
• /
• 제10권2호
• /
• pp.252-259
• /
• 2022

Artificial intelligence is used in fusion with image processing techniques using cameras. Image processing technology is a technology that processes objects in an image received from a camera in real time, and is used in various fields such as security monitoring and medical image analysis. If such image processing reduces the accuracy of recognition, providing incorrect information to medical image analysis, security monitoring, etc. may cause serious problems. Therefore, this paper uses a mixture of YOLOv4-tiny model and image processing algorithm and uses the COCO dataset for learning. The image processing algorithm performs five image processing methods such as normalization, Gaussian distribution, Otsu algorithm, equalization, and gradient operation. For RGB images, three image processing methods are performed: equalization, Gaussian blur, and gamma correction proceed. Among the nine algorithms applied in this paper, the Equalization and Gaussian Blur model showed the highest object detection accuracy of 96%, and the gamma correction (RGB environment) model showed the highest object detection rate of 89% outdoors (daytime). The image binarization model showed the highest object detection rate at 89% outdoors (night).

• Current Optics and Photonics
• /
• 제7권6호
• /
• pp.701-707
• /
• 2023

In laser imaging, accurate extraction of the laser's center is essential. Several methods exist to extract the laser's center in an image, such as the geometric mean, the parabolic curve fitting, and the Gaussian curve fitting, etc. The Gaussian curve fitting is the most suitable because it is based on the physical properties of the laser. The width of the Gaussian laser beam depends on the distance from the laser source to the target object. It is assumed in general that the distance remains constant at a laser spot resulting in a symmetric Gaussian model for the laser image. However, on a curved surface of the object, the distance is not constant; The laser beam is narrower on the side closer to the focal point of the laser light and wider on the side closer to the laser source, which causes the distribution of the laser beam to skew. This study presents a modified Gaussian model in the laser imaging to incorporate the slant angle of a curved object. The proposed method is verified with simulation and experiments.

• Communications for Statistical Applications and Methods
• /
• 제22권6호
• /
• pp.531-542
• /
• 2015

In this survey, we use the normal linear model to demonstrate the use of the linear Bayes method. The superiorities of linear Bayes estimator (LBE) over the classical UMVUE and MLE are established in terms of the mean squared error matrix (MSEM) criterion. Compared with the usual Bayes estimator (obtained by the MCMC method) the proposed LBE is simple and easy to use with numerical results presented to illustrate its performance. We also examine the applications of linear Bayes method to some other distributions including two-parameter exponential family, uniform distribution and inverse Gaussian distribution, and finally make some remarks.

• Journal of Welding and Joining
• /
• 제13권3호
• /
• pp.116-124
• /
• 1995

This paper presents an analytical solution to predict the transient temperature distribution in fillet arc welding including the effect of molten metal. The solution is obtained by solving a transient three-dimensional heat conduction equation with convection boundary conditions on the surfaces of a plate, and mapping the infinite plate onto the fillet weld geometry with energy equation. The electric heat input on the fillet weld and on the infinite plate is assumed to have a combination of two bivariate Gaussian distribution. To check the validity of the solution. FCA welding experiments were performed under various welding conditions. The actual isotherms of the weldment cross-sections at various distances from the arc start point are compared with those of simulation result.

• Journal of Magnetics
• /
• 제2권3호
• /
• pp.105-109
• /
• 1997

The Preisach model neds a density function or Everett function for the hysterisis operator to simulate the hysteresis phenomena. To obtain the function, many experimental data for the first order transition curves are required. However, it needs so much efforts to measure the curves, especially for the hard magnetic materials. By the way, it is well known that the density function has the Gaussian distribution for the interaction axis on the Preisach plane. In this paper, we propose a simple technique to determine the distribution function or Everett function analytically. The initial magnetization curve is used for the distribution of the Everett function for the coercivity axis. A major, minor loop and the initial curve are used to get the Everett function for the interaction axis using the Gaussian distribution function and acceptable results were obtained.

• 천문학회지
• /
• 제42권4호
• /
• pp.71-79
• /
• 2009

In this paper, statistical distribution functions are developed for distance determination of stellar groups. This method depends on the assumption that absolute magnitudes and apparent magnitudes follow a Gaussian distribution function. Due to the limits of the integrands of the frequency function of apparent and absolute magnitudes, we introduce Case A, B, and C Gaussian distributions. The developed approaches have been implemented to determine distances to some clusters and stellar associations. The comparison with the distances derived by different authors reveals good agreement.