• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.835-838
• /
• 2008

In this paper, we present automatic face detection and tracking which is robustness in rotation and translation. Detecting a face image, we used Haar-like feature, which is fast detect facial image. Also tracking, we applied Lucas-Kanade feature tracker and KLT algorithm, which has robustness for rotated facial image. In experiment result, we confirmed that face detection and tracking which is robustness in rotation and translation.

• Natural Product Sciences
• /
• v.16 no.4
• /
• pp.228-232
• /
• 2010

Bioactivity-guided fractionation using on-line HPLC biochemical detection system on $CHCl_3$-soluble fraction of Lycoris radiata led to the isolation of deoxylycorenine (1), O-demethylhomolycorine (2), galanthamine (3), lycoramine (4), mixture of $6{\alpha}$-and $6{\beta}$-haemanthidine (5), and lycorine (6), identified by spectroscopic data and physicochemical property. Among the isolated compounds, 1, 3 and 6 showed acetylcholinesterase inhibitiory activities with $IC_{50}$ values of 18.0, 12.0 and $16.6\;{\mu}M$, respectively, in in vitro colorimetric microplate assay.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.415-418
• /
• 1996

Etching Process, one of the most important process in semiconductor fabrication, has input control part of which components are pressure, gas flow, RF power and etc., and plasma gas which is complex and not exactly understood is used to etch wafer in etching chamber. So this process has not real-time feedback controller based on input-output relation, then it uses EPD(End Point Detection) signal to determine when to start or when to stop etching. Various type EPD controller control etching process using EPD signal obtained from optical intensity of etching chamber. In development EPD controller we concentrate on compensation of this signal intensity and setting the relative signal magnitude at first of etching. We compensate signal intensity using neural network learning method and set the relative signal magnitude using fuzzy inference method. Potential of this method which improves EPD system capability is proved by experiences.

• Journal of Mechanical Science and Technology
• /
• v.15 no.2
• /
• pp.232-238
• /
• 2001

Fast Fourier Transformation is employed to convert the head variation of a pipeline in the time domain to the amplitude of the frequency domain. Applying method of characteristics to a pipeline provides a significant frequency range for a surge introduced from the valve modulation. Inverse Fast Fourier Transformation and a Finite Impulse Response Filter can be used to remove any possible noise existing from the significant frequency range of an unsteady condition. A filtered signal shows higher potential for the inverse calculation of leakage detection than the noise-added signal does. The respective performances of Inverse Fast Fourier Transformation and a Finite Impulse Response Filter are compared in terms of leakage detection capability. Characteristics of the frequency range for multiple leakages were investigated to validate the effectiveness of the noise control method in the frequency domain.

• Proceedings of the SAREK Conference
• /
• 2008.06a
• /
• pp.36-41
• /
• 2008

Fault detection and diagnosis(FDD) system is beneficial in equipment management by providing the operator with tools which can help find out a failure of the system. An experimental study has been performed on fault detection and diagnosis method for a water chiller. Bayes classifier, which is one of classical pattern classifiers, is adopted in deciding whether fault occurred or not. FDD algorithm can detect refrigerant leak failure, when 20% amount of charged refrigerant for normal operation leaks from the water chiller. The refrigerant leak failure caused COP reduction by 6.7% compared with normal operation performance. When two kinds of faults, such as a decrease in the mass flow rate of cooling water and temperature sensor fault of cooling water inlet, are detected, COP is a little decreased by these faults.

• Nuclear Engineering and Technology
• /
• v.36 no.1
• /
• pp.73-82
• /
• 2004

The deaerator of a power plant is one of feedwater heaters in the secondary system, and it is located above the feedwater pumps. The feedwater pumps take the water from the deaerator storage tank, and the net positive suction head(NSPH) should always be ensured. To secure the sufficient NPSH, the deaerator tank is equipped with the level control system of which level sensors are critical items. And it is necessary to ascertain the sensor state on-line. For this, a model-based fault detection and diagnosis(FDD) is introduced in this study. The dynamic control model is formulated from the relation of input-output flow rates and liquid-level of the deaerator storage tank. Then an adaptive state estimator is designed for the fault detection and diagnosis of sensors. The performance and effectiveness of the proposed FDD scheme are evaluated by applying the operation data of Yonggwang Units 3 & 4.

• Bulletin of the Korean Chemical Society
• /
• v.17 no.2
• /
• pp.143-146
• /
• 1996

A continuous potential pulse is applied to a silver-working electrode on a pulsed amperometric detector (PAD) for detection of free cyanide and sulfide. The moving phase is 0.1 M sodium hydroxide, 0.5 M sodium acetate and 5% (v/v) ethylenediamine mixture, and the flow rate is 0.7 mL/min. Optimized pulse conditions include a -200 mV (vs. Ag/AgCl reference electrode) detection potential(Ed) for 60 msec and 50 mV cleaning potential (Ec) for 120 msec. The silver working electrode surface is not poisoned by cyanide or sulfide, and the PAD maintains long-term stability without loss of sensitivity and reproducibility at these pulse conditions. The detection limit of cyanide and sulfide separated by ion chromatography using an anion exchange column is 0.1 ppm and 0.05 ppm, respectively.

• ETRI Journal
• /
• v.46 no.3
• /
• pp.513-525
• /
• 2024

Deep neural networks trained on labeled medical data face major challenges owing to the economic costs of data acquisition through expensive medical imaging devices, expert labor for data annotation, and large datasets to achieve optimal model performance. The heterogeneity of diseases, such as Alzheimer's disease, further complicates deep learning because the test cases may substantially differ from the training data, possibly increasing the rate of false positives. We propose a reconstruction-based self-supervised anomaly detection model to overcome these challenges. It has a dual-subnetwork encoder that enhances feature encoding augmented by skip connections to the decoder for improving the gradient flow. The novel encoder captures local and global features to improve image reconstruction. In addition, we introduce an entropy-based image conversion method. Extensive evaluations show that the proposed model outperforms benchmark models in anomaly detection and classification using an encoder. The supervised and unsupervised models show improved performances when trained with data preprocessed using the proposed image conversion method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.68.4-68
• /
• 2002

$\textbullet$ This report presents an implementation a set of reusable software components which use of fault-tolerance embedded controller for railway signalling systems. These components can be used in real-time applications without application reprogramming. $\textbullet$ This library runs under VxWorks operating system and is oriented on real-time embedded systems. The library includes fault detection, fault containment, checkpointing and recovery components. $\textbullet$ The library enables to support high-speed response to fault occurrence in application software. Garbage collector together with VxWorks Watchdog provides both dead tasks detection and useless resources removing to avoid an overflow. Control flow...

• Journal of Korean Society of Transportation
• /
• v.24 no.3 s.89
• /
• pp.39-50
• /
• 2006

This study aims to introduce a basic principle to improve the incident detection algorithm using traffic flow diagrams that can classify traffic states with a high reliability on the basis of the analysis of traffic flow characteristics under the recurrent or incident congestions. It is tried to newly classify the traffic states with the speed-flow and speed-occupancy diagrams. This is because McMaster algorithm has a tendancy on not identifying the traffic states exactly using the flow-occupancy diagram. In this study it shows that the classification of traffic states is applicable to use speed-occupancy relationship Therefore, it is necessary to determine some parameters to correctly classify the areas representing the traffic states and it may be possible to develop a new algorithm to detect the incident with a high reliability.