• Journal of Korea Multimedia Society
• /
• v.14 no.7
• /
• pp.884-894
• /
• 2011

We proposed a new algorithm for detection of first(S1) and second heart sound(S2). Many researches for detecting primary components and those algorithms have good performance at normal heart sound, but the performance is degraded at abnormal heart sound which is contain murmurs generated by heart disease. Therefore we proposed the S1, S2 detection algorithm using three-order Shannon energy difference. Using S1, S2's character which has large energy difference than murmurs, it is reduced noise and detected S1, S2. According to simulation results, not only normal heart sound but also abnormal heart sound, the proposed algorithm has better performance than former study at abnormal heart sound.

• Journal of the Institute of Convergence Signal Processing
• /
• v.20 no.4
• /
• pp.193-198
• /
• 2019

The wide spread of genuine and fake information on internet has lead to various studies on text analysis. Copying and pasting others' work without acknowledgement, research results manipulation without proof has been trending for a while in the era of data science. Various tools have been developed to reduce, combat and possibly eradicate plagiarism in various research fields. Text similarity measurements can be manually done by using both parametric and non parametric methods of which this study implements cosine similarity and Pearson correlation as parametric while Spearman correlation as non parametric. Cosine similarity and Pearson correlation metrics have achieved highest coefficients of similarity while Spearman shown low similarity coefficients. We recommend the use of non parametric methods in measuring text similarity due to their non normality assumption as opposed to the parametric methods which relies on normality assumptions and biasness.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.12
• /
• pp.2671-2676
• /
• 2009

Empirical mode decomposition (EMD) is new time-frequency analysis method to decompose the signal adaptively and efficiently. The key idea of EMD is to decompose the signal into a set of functions defined by the signal itself, named Intrinsic Mode Functions (IMFs), which preserve the inherent properties of the original signal. Since the decomposition is based on the local time scale of the signal, it is not only applicable to nonlinear and non-stationary processes but also useful in biomedical signals like electrocardiogram (ECG). Traditional low-pass filter uses fourier transform to analysis signal in frequency domain, but EMD is filtered to maintain signal properties in time domain. This paper performed signal decomposition and filtering for noisy ECGs using EMD method. The proposed method is presented and compared with traditional low-pass filter by two performance indices. Our results show effectiveness for enhancement of the noisy ECG waveforms.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2012.05a
• /
• pp.555-561
• /
• 2012

The Hilbert Huang Transform (HHT) technigue with Empirical Mode Decomposition (EMD) is one of the time-frequency domain analysis methods and it has several advantages such that analyzing non-stationary and nonlinear signal is possible. However, there are shortcomings in detecting near-range of frequencies and added noise signals. In this paper, to analyze characteristics of each method, HHT and Short-Time Fourier Transform (STFT) effective in dealing with stationary signals are compared. And with thermoacoustic instabilities signals from a Rijke tube test, HHT and the improved HHT with Ensemble Empirical Mode Decomposition (EEMD) are compared. The results show that the improved HHT is more appropriate than the original HHT due to the relative insensitivity to noise. Therefore it will result in more accurate analysis.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.7
• /
• pp.29-36
• /
• 2021

In this paper, we classify ECG signal data for mobile devices using deep learning models. To classify abnormal heartbeats with high accuracy, three factors of the deep learning model are selected, and the classification accuracy is compared according to the changes in the conditions of the factors. We apply a CNN model that can self-extract features of ECG data and compare the performance of a total of 48 combinations by combining conditions of the depth of model, optimization method, and activation functions that compose the model. Deriving the combination of conditions with the highest accuracy, we obtained the highest classification accuracy of 97.88% when we applied 19 convolutional layers, an optimization method SGD, and an activation function Mish. In this experiment, we confirmed the suitability of feature extraction and abnormal beat detection of 1-channel ECG signals using CNN.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.5
• /
• pp.91-97
• /
• 2016

This paper evaluates the equalization algorithm of FC-MMA (Fast Convergence-Multi Modulus Algorithm) and RMMA (Region based Multi Modulus Algorithm) for the compensation of intersymbol interference which is due to the distortion of communication channel. In order to obtain the error signal for adaptive equalization, the FC-MMA use the modified dispersion constant considering the number of signal symbol, the RMMA separates the 4 region which every symbol points are located, then reduce the symbol point based on this region into constant modulus symbol point. By applying the different principle in order to get the error signal for the updating the tap coefficient of adaptive equalizer, it has the different equalization performance by these error signal. The computer simulation was performed in order to compare the different equalization performance in this paper. The performance index includes the output signal constellation, the residual isi and maximum distortion that is for the convergence characteristics, the SER. As a result of computer simulation, RMMA has more good performance in the residual isi, maximum distortion after in steady state and SER performance than FC-MMA, but not in convergence speed to reach the steady state.

• JOURNAL OF ELECTRICAL WORLD
• /
• s.261
• /
• pp.70-76
• /
• 1998

철도분야에서의 전기$\cdot$신호설비 고장은 열차의 정상운행에 차질을 일으켜 사회적으로 커다란 영향을 끼치게 되므로 설비의 유지관리를 위해서는 많은 노력이 필요하다. 또 이러한 보전작업에는 철도 설비가 널리 산재되어 있기 때문에 이동을 수반하는 작업, 열차가 통과하지 않는 야간 작업, 만일의 경우 위험을 수반하는 작업 등, 이 분야 특유의 작업이 있다. 한편 보전작업에 종사하는 기술자는 해마다 줄어들고 있으며 앞으로 젊은 근로자의 부족은 더욱더 심해질 것으로 보인다. 그러한 이유로 설비보전의 자력화, 미연의 고장방지, 고장의 조기복구와 재발방지를 목표로 지금까지 사람의 손에 의지해 오던 설비보전을 기계화하고 자동화하는 시스템을 개발하게 되었다. 이 시스템은 대상설비에 따라 다음의 3가지 시스템으로 분류된다. $\cdot$신호설비 보전시스템 $\cdot$변전$\cdot$수배전설비 보전시스템$\cdot$연선전기설비 보전시스템 이들시스템에 공통적인 주요 특징을 들면 아래와 같다. (1)사고를 미연에 방지하고 사고원인을 규명 이벤트발생 전후의 변화나 중장기트렌드를 원격감시할 수 있다. (2)시스템의 변경$\cdot$증설이 용이 인텔리전트단말에 의하여 처리를 계층화하였기 때문에 빌딩블록식으로 시스템을 확장할 수 있다. (3)경제적인 시스템 필드 네트워크의 채용으로 공사비를 삭감할 수 있다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.9
• /
• pp.2237-2242
• /
• 2014

As Single households increases for reason of communication development, extending human life, there are many problems occuring all over the world. In order to solve this problem with an invasion of privacy and manintain a healthy life, this paper suggest non-contact type bio-signal measurement system using IR-Radar, displacement sensor and Pan-Tilt system. The proposed system can increse the distance of measured respiration from 1m to over 8m, which is comprised of two IR-Radar for location tracking, one displacement sensor for non-contact type bio-signal measurement and one stepping motor drive system. The proposed system is verified through experiments and were confirmed the possibility.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06a
• /
• pp.469-472
• /
• 2011

신체 부위를 움직이는 상상을 할 때, 일반적으로 뇌의 감각 및 운동 피질 영역에서 특정 주파수 대역의 EEG(Electroencephalography) 신호의 세기가 감소하거나 증가하는 ERD(Event-Related Desynchronization)/ERS(Event-Related Synchronization) 현상이 발생한다. 하지만 ERD/ERS는 현상은 피험자에 의존적이고 매시도마다 큰 차이를 보인다. 이러한 문제를 해결하기 위해, 본 논문에서 각 시간-주파수 공간에 대하여 서로 다른 공간 필터를 구성하는 비 동질(non-homogeneous) 공간 필터 최적화 방법을 제안한다. EEG 신호는 시간에 대하여 비정상적(non-stationary) 특징을 가지기 때문에 제안하는 방법과 같이 시간에 따라 변화하는 ERD/ERS 특징을 반영하여 공간적 특징을 추출하는 방법은 시간에 대한 변화를 고려하지 않은 기존의 방법보다 우수한 성능을 보인다. 본 논문에서는 International BCI Competition IV에서 제공하는 4가지 동작 상상(왼손, 오른손, 발, 혀)에 대한 EEG 신호 데이터를 사용하여 동작 상상 분류 실험을 하고 이 결과를 기존의 타 방법들과 비교 분석하였다. 실험 결과, 피험자에 따라 서로 다른 시간-주파수 특징이 추출됨을 확인하였고, 최적화된 공간 필터들이 시간에 따라 변화하는 것을 확인하였다. 또한 이러한 특징을 이용하여 분류를 수행하였을 때, 더욱 우수한 분류 결과를 보임을 확인하였다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.11a
• /
• pp.388-390
• /
• 2006

본 논문은 종합주가지수, 코스닥 지수의 시계열 일간 데이터의 위상분석을 통해 시계열간의 연관성을 분석하였다. 시계열의 데이터는 비선형, 비정상이다. 따라서 위상성분의 정확한 추출을 위해서 전통적인 수학적 방법이 아닌 순간 위상값을 이용한 새로운 신호분석 방법을 사용하여 두 시계열의 연도별 위상차의 왜도와 첨도값을 기준으로 시계열의 상관특성을 살펴보았다.