• The Journal of the Convergence on Culture Technology
• /
• v.6 no.1
• /
• pp.449-454
• /
• 2020

We propose an improved de-interlacing method that converts the interlaced images into the progressive images by one field. In the first, Inter-pixel values are calculated by applying Newton's forward difference, backward difference interpolation from upper and lower 5 pixel values. Using inter-pixel values obtained from upper and lower 5 pixel values, it makes more accurate a direction estimate by applying the correlation between upper and lower pixel. If an edge direction is determined from the correlation, a missing pixel value is calculated into the average of upper and lower pixel obtained from predicted direction of edge. From simulation results, it is shown that the proposed method improves subjective image quality at edge region and objective image quality at 0.2~0.3dB as quantitative calculation result of PSNR, compared to previous various de-interlacing methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.6
• /
• pp.1-11
• /
• 1998

In this work, we propose a modeling method based on an artifical intelligence technique for a stem generator in a nuclear power plant. Modeling the steam generator is known to be difficult due to several facts; especially, the dynamics of the steam generator is nonminimum phase which is mainly caused by the swell and shrink phenomena from thermal effects. In order to overcome this difficulty, we adopt so-called logic processor whose structure itself has a logical meaning to be easily established and also efficiently learned. Such a manner, we could derive an useful model simulating the dynamics of the steam generator in a nuclear power plant.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.11
• /
• pp.891-895
• /
• 2001

An alternative inverse feedback structure for adaptive active control of periodic noise is introduced for systems with nonminimum phase cancellation path. To obtain the inverse model of the nonminimum phase cancellation path, the cancellation path model can be factorized into a minimum phase term and a maximum phase term. The maximum phase term containing unstable zeros makes the inverse model unstable. To avoid the instability, we alter the inverse model of the maximum phase system into an anti-causal FIR one. An LMS predictor estimates the future samples of the noise, which are necessary for causality of both anti-causal FIR approximation for the stable inverse of the maximum phase system and time-delay existing in the cancellation path. The proposed method has a faster convergence behavior and a better transient response than the conventional filtered-x LMS algorithms with the same internal model control structure since a filtered reference signal is not required. We compare the proposed methods with the conventional methods through simulation studies.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.37 no.11
• /
• pp.720-726
• /
• 1988

This paper deals with an applicaton of some adaptive algorithms to a heat exchanger using a microcomputer and reviews the experimental results obtained. The heat exchanger prepared for experments was identified as a non-minimum phase system and its exact mathematical models was hardly obtainable with direct computation. Thus, classical strategies, such as PID, needed many trial and errors to determine parameters of the controllers. Furthermore such strategies could not guarantee good performance when system parameters vary. To overcome these difficulties and improve performance, two adaptive methods applicable to a non-minimum phase system were chosen and put to the test. In this paper the performance of adaptive controllers is compared with that of conventional PID controller. The final objective of this paper is to construct a controller readily applicable to industrial processes using a microprocessor.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.511-513
• /
• 2000

PaP Smear 테스트는 자궁 경부암 진단에 가장 효율적인 방법으로 알려져 있다. 그러나 이 방법은 높은 위 음성률(false negative error, 15~50%)을 나타내고 있다. 이런 큰 오류율은 주로 다량의 세포 검사에 기인하여, 자동화 시스템의 개발이 절실히 요구되고 있다. 본 논문은 자궁 경부암의 특징인 군집을 이루는 암세포를 인식할 수 있는 시스템을 제안한다. 시스템은 두 부분으로 나누어진다. 첫 단계에서는 저 배율(100배)에서 간단한 영상처리와 최소 근접 트리(Minimum Spanning Tree)를 통해 군집을 이루는 세포를 찾는다. 두 번째 단계서는 고 배율(400배)로 확대하여 군집 세포들로부터 여러 가지 특징을 추출한 후 KNN(k-Neighbor) 방법을 통해 인식하는 단계이다. 50개의 영상 (640X 480, RGB True Color 25 개의 100배 영상 , 25개의 400배 영상)이 실험에 사용되었다. 한 영상을 처리하는데 약 3초 (2.984초) 소요되었으며, 이는 region growing(20초)나 split and merge(58초) 방법 보다 덜 소요되었다. 100배 영상에서 정상과 비정상의 두 그룹으로 나누었을 경우에는 96%의 높은 인식율을 나타내었으나 비정상을 다시 5개의 그룹으로 나누었을 때는 45%로 나타내었다. 이는 영역 추출(segmentation) 단계에서 오류와 트레이닝 데이터의 비정확성에 기인한다. 400배 영상에서는 각각 92%와 30%로 나타내었다. 이는 영역추출 단계에서 사용한 Watershed 방법의 오류로 기인한 것으로 본다.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.9
• /
• pp.1499-1507
• /
• 2003

Physical importance of cutting temperatures has long been recognized. Cutting temperatures have strongly influenced both the tool life and the metallurgical state of machined surfaces. Temperatures in drilling processes are particularly important, because chips remain in contact with the tool for a relatively long time in a hole. Tool temperatures tend to be higher in drilling processes than in other in machining processes. This paper concerns with modeling of thermal behaviors in drilling processes as well as estimation of the cutting temperature distribution based on remote temperature measurements. One- and two-dimensional estimation problems are proposed to analyze drilling temperatures. The proposed thermal models are compared with solutions of finite element methods. Observer algorithms are developed to solve inverse heat conduction problems. In order to apply the estimation of cutting temperatures, approximation methods are proposed by using the solution of the finite element method. In two-dimensional analysis, a moving heat source according to feedrate of the drilling process is regarded as a fixed heat source with respect to the drilling location. Simulation results confirm the application of the proposed methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.4
• /
• pp.996-1007
• /
• 1996

In this paper, a blind adaptation method for a decision feedback equalizer (DFE) is proposed to deal with nominimum phase channels. This equalizer is composed of a linear transversal filter and a prediction error filter which are trained separately using constant modulus and decision feedback prediction algorithms, respectively, during the learnign time. The proposed algorithm guaranetees the DFE to converge to a suboptimal point on the condition that a linear transversal of the proposed scheme is illustrated and the performance is compared with conventional blind equlization algorithms.

• Journal of KIISE:Software and Applications
• /
• v.33 no.12
• /
• pp.999-1007
• /
• 2006

The Backpropagation learning algorithm is known to have slow and false convergence aroused from plateau and local minima. Many substitutes for backpropagation announced so far appear to pay some trade-off for convergence speed and stability of convergence according to parameters. Here, a new algorithm is proposed, which avoids some of those problems associated with the conventional backpropagation problems, especially with local minima, and gives relatively stable and fast convergence with low storage requirement. This is the separate learning algorithm in which the upper connections, hidden-to-output, and the lower connections, input-to-hidden, separately trained. This algorithm requires less computational work than the conventional backpropagation and other improved algorithms. It is shown in various classification problems to be relatively reliable on the overall performance.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.422-425
• /
• 1997

We consider the robust control problem for nonminimum phase(NMP) systems with parametric uncertainty which appear often in aircraft and missile control. First, a new method that makes such an uncertain NMP system to be factored as a interval minimum phase(MP) transfer function and a time delay term in the Pade approximation form has been presented. The controller to be proposed consists of a compensator $C_{Q}$(s) with Smith predictor in the internal model control(IMC) structure, so that it can have good robustness and performance against the structured uncertainty and the time delay behaviour due to NMP plant the $C_{Q}$(s) is designed on the MP model by using QFT. The stability and performance of overall system has been evaluated by the generalized Kharitonov theorem.rem.

• Computational Structural Engineering
• /
• v.5 no.3
• /
• pp.52-55
• /
• 1992

이 글에서는 Mathematica의 고유한 특징과 많은 기능 중의 일부분만을 예를 들어 설명을 하였다. 그러나 Mathematica의 피할 수 없는 단점은 많은 수학적 기능이 포함되어 있기 때문에 처리속도가 늦다는 점이다. 예를 들면 많은 량의 반복작업이나 차수가 큰 매트릭스의 연산작업은 다소 속도가 늦어 PC기종에서는 곤란을 겪을 때가 많다. 따라서 PC대신 workstation 같은 상위기종의 컴퓨터를 이용한다면 처리속도가 빨라져 진행에 문제점이 없다. 한 예로 workstation에서는 차수가 30개인 고유치 해석도 내장함수인 명령어만으로 단지 몇초만에 할 수 있는 데 비하여 PC에서는 기종에 따라 몇배, 몇백배의 시간이 요구되는 것이다. 그리고 또 하나의 단점으로는 방대한 프로그램을 운용하기 위한 비용(ram)이 많이 든다는 점이다. 한 예로 PC에서는 기본적으로 Mathematica를 작동하기 위해 최소한 4 mega ram이 필요하며 여러 수학적 기능을 충분히 이용하기 위해 많은 량의 ram이 필요하다는 점이다. 그러나 위의 단점은 Mathematica가 지니고 있는 고유한 장점을 생각한다면 매우 미미한 것이라 여겨진다. 수학의 대부분의 기능을 포함하고 있으며 기호처리가 가능하고 프로그래밍 기법이 다양하기 때문에 수학을 이용하여 연구를 하는 사람에게는 훌륭한 도구가 생긴 것이라 할 수 있다.