• Journal of Korea Society of Industrial Information Systems
• /
• v.7 no.3
• /
• pp.23-31
• /
• 2002

In this paper we propose a new feature vector for recognition of the facial expression based on Gibbs distributions which are well suited for representing the spatial continuity. The extracted feature vectors are invariant under translation rotation, and scale of an facial expression imege. The Algorithm for recognition of a facial expression contains two parts： the extraction of feature vector and the recognition process. The extraction of feature vector are comprised of modified 2-D conditional moments based on estimated Gibbs distribution for an facial image. In the facial expression recognition phase, we use discrete left-right HMM which is widely used in pattern recognition. In order to evaluate the performance of the proposed scheme, experiments for recognition of four universal expression (anger, fear, happiness, surprise) was conducted with facial image sequences on Workstation. Experiment results reveal that the proposed scheme has high recognition rate over 95%.

• Journal of Korean Society for Atmospheric Environment
• /
• v.24 no.3
• /
• pp.357-366
• /
• 2008

In this study, the analytical performance of trimethylamine (TMA) were investigated with respect to headspace-solid phase microextraction (HS-SPME) method. In order to induce the elution of aqueous TMA to headspace, NaOH was added as a decomposition reagent to aqueous TMA standard. By controlling the combination of three major variables for TMA extraction, the extent of extraction was compared between the two contrasting conditions for each variable (i.e., reaction time (long (L) vs short (S)), exposure temperature (30 vs $50^{\circ}C$), and exposure time (10 vs 30 min)). The results of this comparative analysis showed that the extraction efficiency for all eight types of HS-SPME combinations decreased on the order: L-30-30>L-50-10>L-30-10>L-50-30>S-30-30>S-50-30>S-50-10>S-30-10. The effect of reaction time appeared to exert significant influences on the relative recovery rate of HS-SPME at 90% confidence level. However, the effects of exposure temperature or exposure time were not so significant as reaction time. When the recovery rate of HS-SPME is compared against the direct injection of liquid standard into GC injector, it recorded as 2%. According to this comparative study, the reaction conditions for HS-SPME application can exert significant influences on the analysis of TMA.

• Journal of ILASS-Korea
• /
• v.12 no.1
• /
• pp.18-29
• /
• 2007

Feedwater heaters of many nuclear power plants have recently experienced severe wall thinning damage, which will increase as operating time progresses. Several nuclear power plants in Korea have experienced wall thinning damage in the area around the impingement baffle - installed downstream of the high pressure turbine extraction steam line - inside number 5A and 5B feedwater heaters. At that point, the extracted steam from the high pressure turbine is two phase fluid at high temperature, high pressure, and high speed. Since it flows in reverse direction after impinging the impingement baffle, the shell wall of the number 5 high pressure feedwater heater may be affected by flow-accelerated corrosion. This paper describes the comparisons between the numerical analysis results using the FLUENT code and the down scale experimental data in an effort to determine root causes of the shell wall thinning of the high pressure feedwater heaters. The numerical analysis and experimental data were also confirmed by actual wall thickness measured by an ultrasonic test.

• International Journal of Computer Science & Network Security
• /
• v.24 no.1
• /
• pp.119-126
• /
• 2024

Disease caused by the coronavirus (COVID-19) is sweeping the globe. There are numerous methods for identifying this disease using a chest imaging. Computerized Tomography (CT) chest scans are used in this study to detect COVID-19 disease using a pretrain Convolutional Neural Network (CNN) ResNet50. This model is based on image dataset taken from two hospitals and used to identify Covid-19 illnesses. The pre-train CNN (ResNet50) architecture was used for feature extraction, and then fully connected layers were used for classification, yielding 97%, 96%, 96%, 96% for accuracy, precision, recall, and F1-score, respectively. When combining the feature extraction techniques with the Back Propagation Neural Network (BPNN), it produced accuracy, precision, recall, and F1-scores of 92.5%, 83%, 92%, and 87.3%. In our suggested approach, we use a preprocessing phase to improve accuracy. The image was enhanced using the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm, which was followed by cropping the image before feature extraction with ResNet50. Finally, a fully connected layer was added for classification, with results of 99.1%, 98.7%, 99%, 98.8% in terms of accuracy, precision, recall, and F1-score.

• Korean Journal of Applied Biomechanics
• /
• v.16 no.4
• /
• pp.115-124
• /
• 2006

This study was conducted to investigate the running motion in the max-velocity phase(150-160m) and the fatigue phase(350-360m) during 400m sprint by performed elementary school athletes. Eighteen elementary school male athletes who achieved at least the 3rd place in the sprint at the Korea Gangwon-Do elementary school track and field meetings during 2004 and 2005 were selected as subjects. The running motions performed by the subjects were recorded using two 8mm high speed cameras at the nominal speed of 100 frames per second. The Direct Linear Transformation technique was adopted from the beginning of filming to the final stage of data extraction. KWON 3D motion analysis package program was used to compute the 3 Dimensional coordinates, smoothing factor in which lowpass filtering method was used and cutoff frequency was 6.0 Hz. The movement patterns during foot touchdown and takeoff for the running stride were related with the biomechanical consideration. Within the limitations of this study it is concluded: In order to increase running velocity, several conditions must be fullfilled at the instant of leg touchdown and takeoff during the fatigue phase(350-360m). First, the body C.O.G(Center of Gravity) height should be raised at the instant of leg touchdown and takeoff during the fatigue phase. Second, the foot contact time should be shortened and the takeoff distance should be increased at the foot takeoff during the fatigue phase. Third, the shank angular velocity with respect to a transverse axis through the center of gravity should be increased during the leg touchdown and takeoff in the fatigue phase. Forth, the active landing style described as clawing the ground with the sole of the foot should be performed during the leg touchdown and takeoff in the fatigue phase) phase. Fifth, In order to increase running velocity in the fatigue phase while taking a slightly greater leg knee angle and body lean angle within the range of the subject's running motion during the fatigue phase would result in greater flight distance.

• The KIPS Transactions:PartB
• /
• v.10B no.3
• /
• pp.257-264
• /
• 2003

In the context of a dynamic trading environment, the ultimate goal of the financial forecasting system is to optimize a specific trading objective. This paper proposes a two-phase (extraction and filtering) stock trading system that aims at maximizing the rates of returns. Extraction of stocks is performed by searching specific time-series patterns described by a combination of values of technical indicators. In the filtering phase, several rules are applied to the extracted sets of stocks to select stocks to be actually traded. The filtering rules are automatically induced from past data. From a large database of daily stock prices, the values of technical indicators are calculated. They are used to make the extraction patterns, and the distributions of the discretization intervals of the values are calculated for both positive and negative data sets. We assumed that the values in the intervals of distinctive distribution may contribute to the prediction of future trend of stocks, so the rules for filtering stocks are automatically induced from the data in those intervals. We show the rates of returns when using our trading system outperform the market average. These results mean rule induction method using distributional differences is useful.

• Proceedings of the SAREK Conference
• /
• 2009.06a
• /
• pp.847-852
• /
• 2009

Feedwater heaters of many nuclear power plants have recently experienced severe wall thinning damange, which will increase as operating time progresses. Several nuclear power plants in Korea have experienced wall thinning damage in the area around the impingement baffle inside feed-water heater installed downstream of the turbine extraction stream line. At that point, the extract steam from the turbine is two phase fluid at high temperature, high pressure, and high speed. Since it flows to reverse direction after impinging the impingement baffle, the shell wall of feedwater heaters may be affected by flow-accelerated corrosion. In this paper, to compare wall thinning range according to change entrance nozzle diameter and position with reference numerical analysis model's wall thinning range, various numerical analysis models applied. In case of changing diameter, four different diameter is applied. And a side of nozzle position, two different position-vertical type and parallel type-is applied. And then this paper describes operation of numerical analysis which is composed similar condition with real feed water heater. In conclusion, this study shows effective design for shall wall thinning by changing nozzle diameter and position.

• Applied Chemistry for Engineering
• /
• v.7 no.4
• /
• pp.743-749
• /
• 1996

A new static contractor using capillary phenomena induced by a highly packed fiber bundle was developed for the solvent extraction. When two immiscible phases being cocurrently and forcedly fed into the packed fiber bundle, the contactor brings about a very large liquid-liquid contact area for mass transfer within a small definite space without any turbulence and drop phenomena. In order to test the characteristics and stability of the static contractor system, continuous extraction experiments of TBP-uranyl ion-nitric acid system were done and compared with the batch extraction experiment of the same chemical system. The performance of the static contractor were the same as that of the ideal batch extractor with the same extraction condition. For the increase of the extraction yield by the contactor, the increase of organic flow rate was required at a fixed aqueous flow rate, and a certain residence time of the aqueous phase flow within the contactor system had to be maintained to meet the performance of the batch system of the same phase ratio. The residence time in the case of TBP-uranyl ion-nitric acid system was about 1.9 minutes. This system was confirmed to be effective and stable enough for purposes of the kinetic study of solvent extraction as well as the mutual separation and purification of ordinary materials because of good reproducibility and the stable and large static liquid-liquid contact area.

• Resources Recycling
• /
• v.25 no.1
• /
• pp.60-67
• /
• 2016

A scale-up test with a continuous solvent extraction and electro-winning system was carried out to separate and recover copper from a synthetic sulfuric acid solution (Cu 10.5 g/L, Co 2.0 g/L, Ni 15.0 g/L, Fe 0.2 g/L). The solution was introduced into mixer-settlers with four stages of extraction and two stages of stripping for continuous countercurrent solvent extraction to separate copper from nickel and cobalt. The loading was carried out using 40% LIX 84-I(v/v) as extractant with a phase ratio of A : O = 1 : 1. Meanwhile, the stripping was undertaken at a phase ratio of A : O = 1 : 1.5 using depleted electrolyte containing 35.0 g/L Cu and 180 g/L $H_2SO_4$ as stripping solution. The extraction and stripping efficiencies were found to be 96.7% and 91.0%, respectively. The copper composition of the stripped solution (pregnant electrolyte) was 50.0 g/L Cu with impurities of 25 ppm nickel, 5 ppm cobalt and 3 ppm iron. In the electro-winning process, copper metal of 99.833 purity was yielded with current efficiency of 98.9% and current density of $1.50A/dm^2$.

• Journal of Korean Society for Atmospheric Environment
• /
• v.30 no.4
• /
• pp.362-377
• /
• 2014

Polycyclic aromatic hydrocarbons (PAHs) have been of particular concern since they are present both in the vapor and particulate phases in ambient air. In this study, a simple method was applied to determine the vapor phase PAHs, and the performance of the new method was evaluated with a conventional method. The simple method was based on adsorption sampling and thermal desorption with GC/MS analysis, which is generally applied to the determination of volatile organic compounds (VOCs) in the air. A combination of Carbotrap (300 mg) and Carbotrap-C (100 mg) sorbents was used as the adsorbent. Target compounds included two rings PAHs such as naphthalene, acenaphthylene, and acenaphthene. Among them, naphthalene was listed as one of the main HAPs together with a number of VOCs in petroleum refining industries in the USA. For comparison purposes, a method based on adsorption sampling and solvent extraction with GC/MS analysis was adopted, which is in principle same as the NIOSH 5515 method. The performance of the adsorption sampling and thermal desorption method was evaluated with respect to repeatabilities, detection limits, linearities, and storage stabilities for target compounds. The analytical repeatabilities of standard samples are all within 20%. Lower detection limits was estimated to be less than 0.1 ppbv. In the results from comparison studies between two methods for real air samples. Although the correlation coefficients were more than 0.9, a systematic difference between the two groups was revealed by the paired t-test (${\alpha}$=0.05). Concentrations of two-rings PAHs determined by adsorption and thermal desorption method consistently higher than those by solvent extraction method. The difference was caused by not only the poor sampling efficiencies of XAD-2 for target PAHs and but also sample losses during the solvent extraction and concentration procedure. This implies that the levels of lower molecular PAHs tend to be underestimated when determined by a conventional PAH method utilizing XAD-2 (and/or PUF) sampling and solvent extraction method. The adsorption sampling and thermal desorption with GC analysis is very simple, rapid, and reliable for lower-molecular weight PAHs. In addition, the method can be used for the measurement of VOCs in the air simultaneously. Therefore, we recommend that the determination of naphthalene, the most volatile PAH, will be better when it is measured by a VOC method instead of a conventional PAH method from a viewpoint of accuracy.