• Korean Journal of Remote Sensing
• /
• v.25 no.5
• /
• pp.455-464
• /
• 2009

Lee(2009) proposed the boundary-adaptive despeckling method using a Bayesian model which is based on the lognormal distribution for image intensity and a Markov random field(MRF) for image texture. This method employs the Point-Jacobian iteration to obtain a maximum a posteriori(MAP) estimate of despeckled imagery. The boundary-adaptive algorithm is designed to use less information from more distant neighbors as the pixel is closer to boundary. It can reduce the possibility to involve the pixel values of adjacent region with different characteristics. The boundary-adaptive scheme was comprehensively evaluated using simulation data and the effectiveness of boundary adaption was proved in Lee(2009). This study, as an extension of Lee(2009), has suggested a modified iteration algorithm of MAP estimation to enhance computational efficiency and to combine classification. The experiment of simulation data shows that the boundary-adaption results in yielding clear boundary as well as reducing error in classification. The boundary-adaptive scheme has also been applied to high resolution Terra-SAR data acquired from the west coast of Youngjong-do, and the results imply that it can improve analytical accuracy in SAR application.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.12
• /
• pp.11-19
• /
• 2013

This paper proposes an algorithm using vibration signals and texture analysis for mechanical fault diagnosis of an induction motor. We analyze characteristics of contrast and pattern of an image converted from vibration signal and extract three texture features using gray-level co-occurrence model(GLCM). Then, the extracted features are used as inputs of a multi-level support vector machine(MLSVM) which utilizes the radial basis function(RBF) kernel function to classify each fault type. In addition, we evaluate the classification performance with varying the parameter from 0.3 to 1.0 for the RBF kernel function of MLSVM, and the proposed algorithm achieved 100% classification accuracy with the parameter of the RBF from 0.3 to 1.0. Moreover, the proposed algorithm achieved about 98% classification accuracy with 15dB and 20dB noise inserted vibration signals.

• The Journal of Bigdata
• /
• v.6 no.1
• /
• pp.197-206
• /
• 2021

GPR data is used for underground exploration. The data gathered are interpreted by experts based on experience as the underground facilities often reflect GPR. In addition, GPR data are different in the noise and characteristics of the data depending on the equipment, environment, etc. This often results in insufficient data with accurate labels. Generally, a large amount of training data have to be obtained to apply CNN models that exhibit high performance in image classification problems. However, due to the characteristics of GPR data, it makes difficult to obtain sufficient data. Finally, this makes neural networks unable to learn based on general supervised learning methods. This paper proposes an image classification method considering data characteristics to ensure that the accuracy of each label is similar. The proposed method is based on semi-supervised learning, and the image is classified using clustering techniques after extracting the feature values of the image from the neural network. This method can be utilized not only when the amount of the labeled data is insufficient, but also when labels that depend on the data are not highly reliable.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1542-1550
• /
• 2019

Legacy studies for classifying arrhythmia have been studied to improve the accuracy of classification, Neural Network, Fuzzy, etc. Deep learning is most frequently used for arrhythmia classification using error backpropagation algorithm by solving the limit of hidden layer number, which is a problem of neural network. In order to apply a deep learning model to an ECG signal, it is necessary to select an optimal model and parameters. In this paper, we propose optimal parameter extraction method based on a deep learning. For this purpose, R-wave is detected in the ECG signal from which noise has been removed, QRS and RR interval segment is modelled. And then, the weights were learned by supervised learning method through deep learning and the model was evaluated by the verification data. The detection and classification rate of R wave and PVC is evaluated through MIT-BIH arrhythmia database. The performance results indicate the average of 99.77% in R wave detection and 97.84% in PVC classification.

• Journal of Intelligence and Information Systems
• /
• v.28 no.4
• /
• pp.179-190
• /
• 2022

Data imbalance refers to a phenomenon in which the number of data in one category is too large or too small compared to another category. Due to this, it has been raised as a major factor that deteriorates performance in machine learning that utilizes classification algorithms. In order to solve the data imbalance problem, various ovrsampling methods for amplifying prime number distribution data have been proposed. Among them, SMOTE is the most representative method. In order to maximize the amplification effect of minority distribution data, various methods have emerged that remove noise included in data (SMOTE-IPF) or enhance only border lines (Borderline SMOTE). This paper proposes a method to ultimately improve classification performance by improving the processing method for anomaly data in the traditional SMOTE method that amplifies minority classification data. The proposed method consistently presented relatively high classification performance compared to the existing methods through experiments.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.2
• /
• pp.15-25
• /
• 2024

For effective analysis of animal ecosystems, technology that can automatically identify the current status of animal habitats is crucial. Specifically, animal sound classification, which identifies species based on their sounds, is gaining great attention where video-based discrimination is impractical. Traditional studies have relied on a single deep learning model to classify animal sounds. However, sounds collected in outdoor settings often include substantial background noise, complicating the task for a single model. In addition, data imbalance among species may lead to biased model training. To address these challenges, in this paper, we propose an animal sound classification scheme that combines predictions from multiple models using Focal Loss, which adjusts penalties based on class data volume. Experiments on public datasets have demonstrated that our scheme can improve recall by up to 22.6% compared to an average of single models.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.3
• /
• pp.144-150
• /
• 2006

Clinicians usually use stethoscopic auscultation for the diagnosis of heart diseases. However, the heart sound signal has varying characteristics due to the noise and/or the conditions of the patients. Also, it is not easy for junior clinicians to find the acoustical differences between different kinds or heart sound signals. which may result in errors in the diagnosis. Thus it will be quite useful for the clinicians to make use of an automatic classification system using signal processing techniques. In this paper, we propose to use hidden Markov models in stead of artificial neural networks which have been conventionally used for the automatic classification of heart sounds. In the experiments classifying heart sound signals. we could see that the proposed methods were quite successful in the classification accuracy.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.19 no.7
• /
• pp.726-735
• /
• 2009

In this paper, we propose an approach to signal model-based fault detection and diagnosis system for induction motors. The current fault detection techniques used in the industry are limit checking techniques, which are simple but cannot predict the types of faults and the initiation of the faults. The system consists of two consecutive processes: fault detection process and fault diagnosis process. In the fault detection process, the system extracts the significant features from sound signals using combination of variance, cross-correlation and wavelet. Consequently, the pattern classification technique is applied to the fault diagnosis process to recognize the system faults based on faulty symptoms. The sounds generated from different kinds of typical motor's faults such as motor unbalance, bearing misalignment and bearing loose are examined. We propose two approaches for fault detection and diagnosis system that are waveletand-variance-based and wavelet-and-crosscorrelation-based approaches. The results of our experiment show more than 95 and 78 percent accuracy for fault classification, respectively.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.5
• /
• pp.7-13
• /
• 2020

This paper focuses on detecting abnormal patterns of respiration of humans. In this study, a contact-based device was used to acquire both normal and abnormal respiration signals. To this end, this paper reports the development of a monitoring system to investigate the respiratory status of humans in a normal environment. This work aims to classify the respiratory status, i.e., normal and abnormal status, quantitatively. The respiration signal is acquired using a contact-based medical device (BIOBPAC), and noise reduction is carried out before classifying the respiratory status. To reduce noise, a mixed filter that combines the Savitzky-Golay filter and Median filter is applied to the acquired respiration signals. The inter-class distance is maximized, and the intra-class distance is minimized. The proposed algorithm is straightforward and can be applied to a practical environment. In addition, the experimental results are provided to substantiate the proposed approach.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.3
• /
• pp.207-218
• /
• 1991

In this paper, the characteristics of MODEM with the low and medium transmission rate which are recommended by the international organization such as CCITT have been studied and the features of the telephone channel have been investigated. Then three kinds of classification algorithm for the modulated signals have been suggested and their performance have been researched through the computer simulation. In case that the channel is ideal, the modulated type such as FSK, PSK and QAM could be detected correctly. Moreover, when the modulated signals are received through the telpehone channel or noisy channel, it could be seen that the performance of the correlation algorithm is superior to that of other two algorithms because of the correlation being robust to the characteristics of telephone chanel and Gasussian noise, although the performances of the two classification algorithm using the peak and phase difference, signal constellation and so on decrease rapidly as the amplitude distortion, delay distortion and Gaussian noise increase.