• Journal of Korea Society of Industrial Information Systems
• /
• v.6 no.2
• /
• pp.1-6
• /
• 2001

This paper proposes a pattern classification method of interval vectors by interval neural networks. The proposed method can be applied to pattern classification where attribute values of each sample are given as interval numbers. First, an architecture of interval neural networks is proposed for dealing with interval input vectors. Next, a learning algorithm is derived from the cost function. a cost function is defined using the interval output from the interval neural network and the corresponding target output. Last, using numerical examples, the proposed approach is illustrated and compared with other approach based on the standard back-propagation neural networks.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2489-2491
• /
• 2003

In this study, we implement the laser pattern controller using function generator IC. Overall system consists of : (1) laser excitation circuit and laser tube, (2) two small mirrors to reflect laser beam on the screen, (3) two small motors for X, Y axis enabling each attached mirrors to rotate, (4) controller for motor control and user interface, (5) system power. We explain the architecture of the system and required theory to implement the system. Finally, experimental results are illustrated to show the performance of the system.

• Journal of the Korea Society of Computer and Information
• /
• v.7 no.3
• /
• pp.49-55
• /
• 2002

This paper proposes speaker recognition algorithm which includes both the pitch parameter and the fuzzy. This study proposes a pitch detection method for the peak and valley pitch detection function by means of comparing spectra which utilizes the transform characteristics between time and frequency. It measures the similarity to the original spectrum while arbitrarily varying the period in the time domain. It heavily weights the error due to the changing characteristics of the phonemes, while it is strong against noise. In this paper, makes reference pattern using membership function and performs vocal track recognition of common character using fuzzy pattern matching in odor to include time variation width for non-linear utterance time.

• Journal of the Korea Society of Computer and Information
• /
• v.5 no.3
• /
• pp.51-57
• /
• 2000

This paper proposes speaker recognition algorithm which includes both the pitch and the fuzzy. This study proposes a pitch detection method for the peak and valley pitch detection function by means of comparing spectra which utilizes the transform characteristics between time and frequency. It measures the similarity to the original spectrum while arbitrarily varying the period in the time domain. It heavily weights the error due to the changing characteristics of the phonemes, while it is strong against noise. In this paper, makes reference pattern using membership function and performs vocal track recognition of common character using fuzzy pattern matching in order to include time variation width for non-linear utterance time.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.107-115
• /
• 2006

This paper presents a new class of activation functions for Cascade Correlation learning algorithm, which herein will be called CosGauss function. This function is a cosine modulated gaussian function. In contrast to the sigmoidal, hyperbolic tangent and gaussian functions, more ridges can be obtained by the CosGauss function. Because of the ridges, it is quickly convergent and improves a pattern recognition speed. Consequently it will be able to improve a learning capability. This function was tested with a Cascade Correlation Network on the two spirals problem and results are compared with those obtained with other activation functions.

• Proceedings of the KIEE Conference
• /
• 1990.11a
• /
• pp.428-431
• /
• 1990

This paper describes a pattern classification algorithm of QRS-complexes using significant point detection for extracting features of signals. Significant point extraction was processed by zero-crossing method, and decision function based on relational spectrum was used for pattern classification of the QRS-complexes. The hierarchical AND/OR graph was obtained by decomposing the signal, and by use of this graph, QRS's patterns were classified. By using the proposed algorithm, the accuracy of pattern classification and the processing speed were improved.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10b
• /
• pp.1737-1741
• /
• 1991

The authors proposed a new two-dimensional(2D) positioning system by use of M-array and correlation technique which is suitable for noisy environment in '88KACC and its revised versions in '89KACC and '90KACC. This system uses the property of M-array that the autocorrelation function of the M-array has a sharp peak at its origin. In this paper, a new mark pattern is developed, instead of M-array, with which the two-dimensional positioning system becomes robust to rotation error of TV camera. The property of the rotation-free pattern is checked under various conditions, and it is shown that, by use of this rotation-free pattern, the positioning system can be used not only in a noisy environment but also in a roughly aligned set up of the TV camera.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2001.05a
• /
• pp.783-788
• /
• 2001

Voids created with concrete construction or deterioration result in serious weakness from the aspects of both structural and durable function. Ultrasonic method using image processing technique was used for detecting pattern of voids in concrete in this study Experimental investigation was carried out for three types(patterns) of void in concrete. The effect of curing period of concrete and ultrasonic measurement method was also investigated. As a result it has been verified that the semi-direct measurement method is more effective than the other methods for detecting pattern of voids in concrete in ultrasonic method using image processing technique. The longer the curing period of concrete is, the better the detection accuracy of void pattern can be obtained.

• Smart Structures and Systems
• /
• v.7 no.2
• /
• pp.117-132
• /
• 2011

This paper is aimed at presenting two methods on the basis of pattern search and genetic algorithms to detect and estimate damage in plates using the modal data of a damaged plate. The proposed methods determine the damages of plate structures using optimization of an objective function by pattern search and genetic algorithms. These methods have been applied to two numerical examples, namely four-fixed supported and cantilever plates with and without noise in the modal data and containing one or several damages. The obtained results clearly reveal that the proposed methods can be viewed as a powerful and reliable method for structural damage detection in plates using the modal data.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.18 no.4
• /
• pp.321-326
• /
• 2005

Chemical mechanical polishing(CMP) process is essential technology to be applied to manufacturing the dielectric layer and metal line in semiconductor devices. It has been known that overpolishing in CMP depends on pattern selectivity as a function of density and pitch, and use of fixed abrasive pad(FAP) is one method which can improve the pattern selectivity. Thus, dishing & erosion defects can be reduced. This paper introduces the manufacturing technique of FAP using hydrophilic polymers with swelling characteristic in water and explains the self-conditioning phenomenon. When applied to tungsten blanket wafers, the FAP resulted in appropriate performance in point of uniformity, material selectivity and roughness. Especially, reduced dishing and erosion was observed in CMP of tungsten pattern wafer with the proposed FAP.