• Journal of Broadcast Engineering
• /
• v.21 no.2
• /
• pp.268-271
• /
• 2016

In video sequences, foreground object detection being composed of a background model and a background subtraction is an important part of diverse computer vision applications. However, object detection might fail in sudden illumination changes. In this letter, an illumination-robust background detection is proposed to address this problem. The method can provide quick adaption to current illumination condition using two background models with different adaption rates. Since the proposed method is a non-parametric approach, experimental results show that the proposed algorithm outperforms several state-of-art non-parametric approaches and provides low computational cost.

• Structural Engineering and Mechanics
• /
• v.61 no.4
• /
• pp.497-510
• /
• 2017

A framed structure may be composed of two sub-structures, which are linked by a hinged joint. One sub-structure is the primary system and the other is the secondary system. The primary system, which is subjected to the periodic external load, can give rise to an auto-parametric resonance of the second system. Considering the geometric-stiffness effect produced by the axially internal force, the element equation of motion is derived by the extended Hamilton's principle. The element equations are then assembled into the global non-homogeneous Mathieu-Hill equations. The Newmark's method is introduced to solve the time-history responses of the non-homogeneous Mathieu-Hill equations. The energy-growth exponent/coefficient (EGE/EGC) and a finite-time Lyapunov exponent (FLE) are proposed for determining the auto-parametric instability boundaries of the structural system. The auto-parametric instabilities are numerically analyzed for the two frames. The influence of relative stiffness between the primary and secondary systems on the auto-parametric instability boundaries is investigated. A phenomenon of the "auto-parametric internal resonance" (the auto-parametric resonance of the second system induced by a normal resonance of the primary system) is predicted through the two numerical examples. The risk of auto-parametric internal resonance is emphasized. An auto-parametric resonance experiment of a ${\Gamma}$-shaped frame is conducted for verifying the theoretical predictions and present calculation method.

• Journal of information and communication convergence engineering
• /
• v.11 no.2
• /
• pp.95-102
• /
• 2013

This paper presents a stochastic driver behavior modeling framework which takes into account both individual and general driving characteristics as one aggregate model. Patterns of individual driving styles are modeled using a Dirichlet process mixture model, as a non-parametric Bayesian approach which automatically selects the optimal number of model components to fit sparse observations of each particular driver's behavior. In addition, general or background driving patterns are also captured with a Gaussian mixture model using a reasonably large amount of development data from several drivers. By combining both probability distributions, the aggregate driver-dependent model can better emphasize driving characteristics of each particular driver, while also backing off to exploit general driving behavior in cases of unseen/unmatched parameter spaces from individual training observations. The proposed driver behavior model was employed to anticipate pedal operation behavior during car-following maneuvers involving several drivers on the road. The experimental results showed advantages of the combined model over the model adaptation approach.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.10C
• /
• pp.1013-1022
• /
• 2003

This paper implemented the real-time object tracking algorithm that extracts and tracks the moving object adaptively to input frame sequence by using non-parametric image processing method and Kalman filter-based dynamic AR(2) process method. By applying non-parametric image processing to input frames, the moving object was extracted from the background adaptively to diverse environmental conditions. And the movement of object was able to be adaptively estimated and tracked by modeling the various movement of object as dynamic AR(2) process and estimating based on the Kalman filter the parameters of AR(2) process dynamically changing along time. The experiments of the implemented object tracking system showed that the proposed method tracked the moving object as more approximately as the estimation error became about l/2.5∼1/50 of one of the traditional tracking method based on linear Kalman filter.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2010.09a
• /
• pp.416-422
• /
• 2010

Estimating condition of geotechnical structures are difficult because of nonlinear time dependency and seasonal effects. Measuring data of structure failure is highly variable in time and space, and a unique approach cannot be defined to model structure movements. Characteristics of movements are obtained by using a statistical method called Principal Component Analysis(PCA). The PCA is a non-parametric method to separate unknown, statistically uncorrelated source processes from observed mixed processes. Instead, since the "best" mathematical relationship is estimated for given data sets of the input and output measured from target systems. As a consequence, this method is advantageous in modeling systems whose geomechanical properties are unknown or difficult to be measured.

• Journal of Applied Reliability
• /
• v.11 no.1
• /
• pp.43-57
• /
• 2011

In the paper, we consider two-dimensional warranty policy with failure times and repair times. The failure times are considered within the warranty period and the repair times are considered within the repair time limit. Under the renewable warranty policy and non-renewable warranty policy, we consider the number of warranty services in the censored area by warranty period and repair time limit to conduct warranty cost analysis. We investigate the field data to check their dependency and implement our proposed approaches to conduct warranty cost analysis using the parametric methods. Numerical examples are discussed to demonstrate the applicability of the methodologies and results based on the proposed approach in the paper.

• Korean Journal of Social Welfare Studies
• /
• no.39
• /
• pp.5-28
• /
• 2008

The purpose of this study is two-fold: 1) to identify different adolescent smoking trajectories in Korea; and 2) to examine predictors of those smoking trajectories within a social developmental frame. Data were from the Korea Youth Panel Survey(KYPS), a longitudinal study of 3,449 youths followed since 2003. Using semi-parametric group-based modeling, four smoking trajectories were identified: non initiators, late onsetters, experimenters, and escalators. Multinomial logistic regressions were then used to identify risk and protective factors that distinguish the trajectory groups from one another. Among non smokers at age 13, late onsetters were distinguished from non initiators by a variety of factors in every ecological domain. Among youths who already smoked at age 13, escalators who increased their smoking were distinguished from experimenters who almost desisted from smoking by age 17 by self-esteem and academic achievement. Finally, implications for youth welfare practice from this study were discussed.

• Biotechnology and Bioprocess Engineering:BBE
• /
• v.11 no.1
• /
• pp.67-71
• /
• 2006

This paper compares regression and neural network modeling approaches to predict competitive biosorption equilibrium data. The regression approach is based on the fitting of modified Langmuir-type isotherm models to experimental data. Neural networks, on the other hand, are non-parametric statistical estimators capable of identifying patterns in data and correlations between input and output. Our results show that the neural network approach outperforms traditional regression-based modeling in correlating and predicting the simultaneous uptake of copper and cadmium by a microbial biosorbent. The neural network is capable of accurately predicting unseen data when provided with limited amounts of data for training. Because neural networks are purely data-driven models, they are more suitable for obtaining accurate predictions than for probing the physical nature of the biosorption process.

• Computers and Concrete
• /
• v.16 no.3
• /
• pp.415-428
• /
• 2015

Study on flexural retrofitting of RC beams using external bars with additional intermediate anchorages at soffit is reported in this paper. Effects of varying number of anchorages in the external bars at soffit were studied by finite element analysis using ANSYS 12.0 software. The results were also compared with available experimental results for beam with only two end anchorages. Two sets of reference and retrofitted beam specimens with two, three, four and five anchorages were analysed and the results are reported. FE modeling and non-linear analysis was carried out by discrete reinforcement modeling using Solid65, Solid45 and Link8 elements. Combin39 spring elements were used for modeling the frictional contact between the soffit and the external bars. The beam specimens were subjected to four-point bending and incremental loading was applied till failure. The entire process of modeling, application of incremental loading and generation of output in text and graphical format were carried out using ANSYS Parametric Design Language.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.05a
• /
• pp.369-373
• /
• 2001

In this paper, we proposes the steps adopted to construct the neural network model for GMAW welds. Conventional, automated process generally involves sophisticated sensing and control techniques applied to various processing parameters. Welding parameters are influenced by numerous factors, such as welding current, arc voltage, torch travel speed, electrode condition and shielding gas type and flow rate etc. In traditional work, the structural mathematical models have been used to represent this relationship. Contrary to the traditional model method, neural network models are based on non-parametric modeling techniques. For the welding process modeling, the non-linearity at well as the coupled input characteristics makes it apparent that the neural network is probably the most suitable candidate for this task. Finally, a suitable proposal to improve the construction of the model has also been presented in the paper.