• Communications for Statistical Applications and Methods
• /
• v.19 no.4
• /
• pp.619-628
• /
• 2012

Noise reduction is an important issue in the field of image processing because image noise lowers the quality of the original pure image. The basic difficulty is that the noise and the signal are not easily distinguished. Simple smoothing is the most basic and important procedure to effectively remove the noise; however, the weakness is that the feature area is simultaneously blurred. In this research, we use ways to measure the degree of noise with respect to the degree of image features and propose a Bayesian noise reduction method based on MAP (maximum a posteriori). Simulation results show that the proposed adaptive noise reduction algorithm using Bayesian MAP provides good performance regardless of the level of noise variance.

• Journal of Institute of Control, Robotics and Systems
• /
• v.12 no.12
• /
• pp.1202-1208
• /
• 2006

Development of a model predictive control(MPC) algorithm and its application to the condensate recycle process of a cogeneration power station has been conducted. The cogeneration power station has different characteristics from other industrial processes where MPC has been dominantly applied in that the operating mode changes continuously with seasons and we Ether. Such a characteristic makes it difficulty, a linearized model was derived from mass and pressure balances and linearization. The MPC algorithm has been developed so that the controller tuning is easy with one tuning knob for each output and the constrained optimization is solved by an interior point method. Performance of the MPC algorithm has been verified with the numerically simulated process under various disturbance scenarios and mode changes.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2007.11a
• /
• pp.415-421
• /
• 2007

This paper proposes a convolution tree kernel-based approach for relation extraction where the parse tree is expanded with entity features such as entity type, subtype, and mention level etc. Our study indicates that not only can our method effectively capture both syntactic structure and entity information of relation instances, but also can avoid the difficulty with tuning the parameters in composite kernels. We also demonstrate that predicate verb information can be used to further improve the performance, though its enhancement is limited. Evaluation on the ACE2004 benchmark corpus shows that our system slightly outperforms both the previous best-reported feature-based and kernel-based systems.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.7
• /
• pp.2390-2406
• /
• 2022

Scene text recognition has important application value and attracted the interest of plenty of researchers. At present, many methods have achieved good results, but most of the existing approaches attempt to improve the performance of scene text recognition from the image level. They have a good effect on reading regular scene texts. However, there are still many obstacles to recognizing text on low-quality images such as curved, occlusion, and blur. This exacerbates the difficulty of feature extraction because the image quality is uneven. In addition, the results of model testing are highly dependent on training data, so there is still room for improvement in scene text recognition methods. In this work, we present a natural scene text recognizer to improve the recognition performance from the feature level, which contains feature representation and feature enhancement. In terms of feature representation, we propose an efficient feature extractor combined with Representative Batch Normalization and ResNet. It reduces the dependence of the model on training data and improves the feature representation ability of different instances. In terms of feature enhancement, we use a feature enhancement network to expand the receptive field of feature maps, so that feature maps contain rich feature information. Enhanced feature representation capability helps to improve the recognition performance of the model. We conducted experiments on 7 benchmarks, which shows that this method is highly competitive in recognizing both regular and irregular texts. The method achieved top1 recognition accuracy on four benchmarks of IC03, IC13, IC15, and SVTP.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.370-370
• /
• 2012

Feature size of Cu interconnects keep shrinking into several tens of nanometer level. For this reason, the Cu interconnects face challenging issues such as increase of electro-migration, line-width dependent electrical resistivity increase, and gap-filling difficulty in high aspect ratio structures. As the thickness of the Cu film decreases below 30 nm, the electrical resistivity is not any more constant, but rather exponential. Research on alloying with other elements have been started to inhibit such escalation in the electrical resistivity. A faint trace of Al added in Cu film by sputtering was reported to contribute to suppression of the increase of the electrical resistivity. From an industrial point of view, we introduced cyclic metal organic chemical vapor deposition (MOCVD) in order to control Al concentration in the Cu film more easily by controlling the delivery time ratio of Cu and Al precursors. The amount of alloying element could be lowered at level of below 1 at%. Process of the alloy formation was applied into gap-filling to evaluate the performance of the gap-filling. Voidless gap-filling even into high aspect ratio trenches was achieved. In-depth analysis will be discussed in detail.

• Journal of the Ergonomics Society of Korea
• /
• v.36 no.3
• /
• pp.197-211
• /
• 2017

Objective: We propose to apply color information to facilitate finding books in the library. Background: Currently, books are classified in the basis of a decimal classification system and a call number in the library. Users find a book using the call number. However, this classification system causes various difficulties. Method: In a process analysis and survey study, we identify what the real problem is and where the problem is occurred. To solve the real problems, we derived a new search method using color information. We conducted a comparative experiment with 48 participants to see whether the new method can show higher performance. Results: The new method using color information showed faster time and higher subjective rating scores than current call number method. Also, the new method showed faster time regardless of the skill level while the call number method showed time differences in terms of the skill level. Conclusion: The effectiveness of the proposed method was verified by experiments. Users will be able to find the desired book without difficulty. This method can improve the quality of service and satisfaction of library use. Application: Our book search method can be applied as a book search tool in a real public library. We hope that the method can provide higher satisfaction to users.

• Journal of the Korean Society for Precision Engineering
• /
• v.19 no.12
• /
• pp.38-44
• /
• 2002

A walking training robot is proposed to provide stable and comfortable walking supports by reducing body weight load partially and a force control of an arm of walking training robot using sliding mode controller is also proposed. The current gait training apparatus in hospital are ineffective for the difficulty in keeping constant unloading level and for the constraint of patients' free walking. The proposed walking training robot effectively unloads body weight during walking. The walking training robot consists of an unloading manipulator and a mobile platform. The manipulator driven by an electro-mechanical linear mechanism unloads body weight in various levels. The mobile platform is wheel type, which allows patients to walt freely. The developed unloading system has advantages such as low noise level, lightweight, low manufacturing cost and low power consumption. A system model fur the manipulator is established using Lagrange's equation. To unload the weight of the patients, sliding mode control with p-control is adopted. Both control responses with a weight and human walking control responses are analyzed through experimental implementation to demonstrate performance characteristics of the proposed force controller.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.12 no.4
• /
• pp.106-113
• /
• 1998

The steam generator waste level control system in a nuclear power station has difficulty in its mathematical modeling and theoretical application in both a transient and steady state operation. Therefore, the stability problem of the conventional control methods brings many researches interests to the various methods of a system design in recent years. In this study, an improved loop shaping approach is proposed by applying the genetic algorithm to QFT (Quantitative Feedback Theory) in designing a control system in order to the performance of the system. And the effects of the proposed methods are shown by the simulation results.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.6
• /
• pp.66-74
• /
• 1994

Simulation of sonar reverberation time series is very useful because most acoustic models are power level models and have a difficulty when performance of hardware system is evaluated under the reverberant condition. Thus, in this paper, the simulation of reverberation time series is attempted, First, normalized spectrum, whose bandwidth is varying in the frequency domain and which has zero-mean Gaussian distribution, is calculated at pre-selected receiving time. Second, reverberation levels given by underwater acoustic model are combined with normalized spectrum in the frequency domain. Finally, nonstationary sonar reverberation time series are simulated by IFT(Inverse Fourier Transform).

• Journal of Korean Society for Atmospheric Environment
• /
• v.15 no.6
• /
• pp.779-789
• /
• 1999

To support daily ground-level $O_3$ forecasting in Seoul, a transfer function model(TFM) has been developed by using surface meteorological data and pollutant data(previous-day [$O_3$] and [$NO_2$]) from 1 May to 31 August in 1997. The forecast performance of the TFM was evaluated by statistical comparison with $O_3$ concentration observed during September it is shown that correlation coefficient(R), root mean squared error(RMSE), normalized mean squared error(NMSE) and mean relative error(MRE) were 0.73, 15.64, 0.006 and 0.101, respectively. The TFM appeared to have some difficulty forecasting very high $O_3$ concentrations. To compare with this model, multiple regression model(MRM) was developed for the same period. According to statistical comparison between the TFM and MRM. two models had similar predictive capability but TFM based on $O_3$ concentration higher than 60 ppb provided more accurate forecast than MRM. It was concluded that statistical model based on TFM can be useful for improving the accuracy of local $O_3$ forecast.