• 반도체디스플레이기술학회지
• /
• 제22권4호
• /
• pp.166-171
• /
• 2023

This paper presents the synaptic characteristics of IGZO memristors in neuromorphic computing, using MATLAB/Simulink and NeuroSim. In order to investigate the variations in the conductivity of IGZO memristor and the corresponding changes in the hidden layer, simulations are conducted by using the MNIST dataset. It was observed from simulation results that the recognition accuracy could be dependent on various parameters of IGZO memristor, along with the experimental exploration. Moreover, we identified optimal parameters to achieve high accuracy, showing an outstanding accuracy of 96.83% in image classification.

• 대한기계학회논문집A
• /
• 제21권5호
• /
• pp.835-841
• /
• 1997

Microscopic examination of the morphology of wear debris is an accepted method for machine condition and fault diagnosis. However wear particle analysis has not been widely accepted in industry because it is dependent on expert interpretation of particle morphology and subjective assessment criteria. This paper was undertaken to analyze the morphology of wear debris for machine condition diagnosis of the lubricated moving surfaces by image processing and analysis. The lubricating wear test was performed under different sliding conditions using a wear test device made in our laboratory and wear testing specimen of the pin-on-disk-type was rubbed in paraffine series base oil. In order to describe characteristics of debris of various shape and size, four shape parameters (50% volumetric diameter, aspect, roundness and reflectivity) have been developed and outlined in the paper. A system using such techniques promises to obviate the need for subjective, human interpretation of particle morphology in machine condition monitoring, thus to overcome many of the difficulties in current methods and to facilitate wider use of wear particle analysis in machine condition monitoring.

• 방송공학회논문지
• /
• 제2권1호
• /
• pp.16-23
• /
• 1997

HDTV의 다음세대 TV로서 주목받고 있는 3D TV의 구현을 위한 주요 기술중 보는이의 시점에 해당하는 영상을 보여주는 다시점 입체영상표시가 중요한 연구대상으로 대두되고 있다. 본 논문에서는 다시점 입체영상표시를 구현하기 위한 기존방법들의 문제점으로 지적되는 많은 연산량 문제를 해결할 수 있는 새로운 알고리즘을 제안한다. 다시점 영상모음을 영상 공간축에서 을로 다운 샘플링(down sampling )하여 다해상도 영상 피라미드를 만들고, 이를 바탕으로 저해상도 EPI에서부터 대응점들을 탐색하고, 그에 해당하는 상위 해상도에서는 하위 해상도에서 탐색한 결과를 확인·교정해 가는 방법을 사용하므로써, 기존의 방법보다 고속처리가 가능하며, 잡음에도 강인한 기법을 제안한다.

• International journal of advanced smart convergence
• /
• 제11권4호
• /
• pp.155-162
• /
• 2022

The purpose of this study is to conduct an empirical study on brand authenticity targeting sportswear brand consumers. Through this, we intend to provide the accumulation and implications of authenticity research. For the research model, first, the authenticity of sportswear brand companies was selected as an independent variable. Brand image and brand attitude were selected as the next parameters. Finally, the dependent variable was the intention to pay the premium. Structural equation model analysis was conducted for the structural relationship between these variables. The subjects of this study are consumers who have purchased sportswear brands within the past year. Convenience sampling was used for the sample survey, and 262 people were finally selected as valid samples. The survey was conducted as a non-face-to-face online survey due to the COVID-19 infection. For data processing, frequency analysis was conducted using SPSS 23 to identify the individual characteristics of the survey subjects. In addition, exploratory factor analysis and reliability analysis were performed to refine the scale of the survey tool. Next, using AMOS 21, confirmatory factor analysis and correlation analysis were conducted to verify the measurement model. In addition, structural equation model analysis was conducted to verify the hypothesis. As a result of the analysis, all six hypotheses selected from the research model were adopted.

• 한국정밀공학회지
• /
• 제19권6호
• /
• pp.136-144
• /
• 2002

Melting method has long been considered difficult to realize because of problems such as the low foamability of molten metal, the varying size of cellular structures and solidification shrinkage. The parameters to solve the problem in electric furnace were stirring temperature, stirring velocity, heating velocity and foaming temperature It is important to consider the effects of induction heating, because it brings about the inner flow by the temperature gradient. Aspect ratio also depends on the induction heating. Mechanical properties are dependent on cell sizes and aspect rations. Therefore, this paper presents the effects of these parameters on the cell sizes. For the sake of this, combined stirring process was used to fabricate aluminum foam materials by the above mentioned parameters. Image analysis was performed to calculate the cell sizes, distributions, and aspect ratioes at the cross section of feared aluminum in the direction of height.

• 한국방사선학회논문지
• /
• 제11권6호
• /
• pp.501-508
• /
• 2017

방사선치료에 사용하는 CT 스캐너에 의해 획득된 CT 및 CBCT 전자밀도팬텀의 CT영상부터 CT수 대 물리적 밀도 변환에 관한 CT 스캐닝 매개변수의 의존성은 실험으로 분석하였다. CT수는 관전류량, 슬라이스 두께, 영상재구성 필터, 시야 그리고 팬텀 용적의 크기에 대해 의존하지 않았다. 그러나 CT수는 관전압과 팬텀 횡단면적 크기에 의존하였다. 결과로서, 물리적 밀도 1이상의 범위에 대하여, 90과 120 kVp 사이의 관전압에서 관측된 최대 CT수 차이는 27%이었고, 그리고 CT 몸통과 머리 전자밀도팬텀 사이에서 관측된 최대 CT수 차이는 15%이었다.

• 대한원격탐사학회지
• /
• 제25권2호
• /
• pp.95-105
• /
• 2009

지상 관측으로부터 수집된 시계열 원격탐사 자료는 관측환경의 악화와 감지 시스템의 기계적 고장과 같은 관측 장애요인에 의해 많은 미관측 및 악성 자료를 가지게 된다. 육상의 지표면 parameters는 기후와 주로 연관되어 있으므로 육상 관측 위성 영상에 나타나는 많은 물리적 과정은 계절 주기에 따른 시간적 변화를 보인다. 본 연구에서 제안된 적응 feedback 시스템은 계절에 따라 변하는 물리적 과정을 포함하는 시계열 원격 탐사 영상 시리즈를 재구축한다. 이 시스템에서는 계절적 변화를 추적하기 위하여 하모닉 모형을 사용하고 수치 영상 모형의 공간적 의존성을 나타내기 위해 Gibbs Random Field를 사용한다. 재구축 과정을 통하여 구성된 적응 하모닉 모형을 사용하여 지표면 연속적 변화를 감시할 수 있다. 본 연구에서는 1996년부터 2000년까지 한반도로부터 관측된 AVHRR 영상 시리즈를 일 주일 간격으로 정적 합성하여 NDVI 시리즈를 구하고 하모닉 모형을 사용하는 적응 재구축 시스템을 이 NDVI 시리즈에 적용하여 한반도 식생 변화를 추적하였다. 연구 결과는 하모닉 적응 재구축 시스템이 실시간 지표면 변화 감시를 하는데 매우 효과적인 수단이 될 것이라는 잠재성을 보여준다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.1162-1164
• /
• 2003

The methods of the geometric reconstruction and sensor calibration of satellite linear pushbroom images are investigated. The model of the sensor used is based on the SPOT model that is developed by Kraiky. The satellite trajectory is a Keplerian trajectory in the approximation. Four orbit parameters, longitude of the ascending node(${\omega}$), inclination of the orbit plan(I), latitude argument of the satellite(W) and distance between earth center and satellite, are used for the camera modeling. Time-dependent orbit parameters are expressed by quadratic polynomials. SPOT-5 images have been used for validation tests. The results are that the RMSE acquired from 20 GCPs is 1.763m and the RMSE of 5 checking points 2.470m. Because the ground resolution of SPOT-5 is 2.5m, the result obtained in this study has a good accuracy. It demonstrates that the sensor model developed by this study can be used to reconstruct the geometry of satellite image using pushbroom camera.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제51권11호
• /
• pp.514-520
• /
• 2002

n industrial computer vision systems, the image quality is dependent on the parameters such as light source, illumination method, optics, and surface properties. Most of them are related with the lighting system, which is designed in heuristic based on the designer's experimental knowledge. In this paper, a design code by which the optimal lighting method and light source for computer vision systems can be found are suggested based on experimental results. The design coed is applied to the design of the lighting system for the transistor marking inspection system, and the overall performance of the machine vision system with the lighting system show the effectiveness of the proposed design code.

• 제어로봇시스템학회논문지
• /
• 제14권5호
• /
• pp.438-443
• /
• 2008

In this paper, segmenting and classifying low contrast defects on flat panel display is one of the key problems for automatic inspection system in practice. Problems become more complicated when the quality of acquired image is degraded by the illumination irregularity. Many algorithms are developed and implemented successfully for the defects segmentation. However, vision algorithms are inherently prone to be dependent on parameters to be set manually. In this paper, one morphological segmentation algorithm is chosen and a technique using frequency domain analysis of input images is developed for automatically selection the morphological parameter. An extensive statistical performance analysis is performed to compare the developed algorithms.