• 한국컴퓨터정보학회논문지
• /
• 제8권2호
• /
• pp.31-36
• /
• 2003

본 논문은 주성분 분석에 의한 특징 탐색 기법의 가속화 방법을 제안하고 있다. 특징 탐색이란 주어진 영상이 찾고자 하는 특징을 포함하고 있는지의 여부를 판단하고, 만일 그 특징이 포함되어 있다면 그 위치와 영역을 결정하는 방법이다. 탐색 대상으로 하는 얼굴 또는 특징의 위치와 스케일을 미리 알 수 없으므로 모든 위치에서 다양한 스케일의 특징에 대한 존재 가능성을 계산해야하는데 이것은 방대한 공간에서의 탐색문제이다. 본 논문에서는 다단계 예측맵과 윤곽선 예측맵을 이용함으로써 탐색공간을 축소하고 빠른 얼굴 및 특징 탐색을 가능케하는 방법을 제안하고 있다. 실험결과, 제안된 방법이 기존의 전역탐색방법에 비하여 계산량을 10％이하로 줄일 수 있었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 D
• /
• pp.3037-3040
• /
• 2000

Real-Time object tracking has emerged as an important component in several application areas including machine vision. surveillance. Human-Computer Interaction. image-based control. and so on. And there has been developed various algorithms for a long time. But in many cases. they have showed limited results under uncontrolled situation such as illumination changes or cluttered background. In this paper. we present a novel. computationally efficient algorithm for tracking human face robustly under illumination changes and cluttered backgrounds. Previous algorithms usually defines color model as a 2D membership function in a color space without consideration for illumination changes. Our new algorithm developed here. however. constructs a 3D color model by analysing plenty of images acquired under various illumination conditions. The algorithm described is applied to a mobile head-eye robot and experimented under various uncontrolled environments. It can track an human face more than 100 frames per second excluding image acquisition time.

• 한국축산식품학회지
• /
• 제42권4호
• /
• pp.672-688
• /
• 2022

The objective of this study was to explore the potential of front face fluorescence spectroscopy (FFFS) as rapid, non-destructive and inclusive technique along with multi-variate analysis for predicting meat adulteration. For this purpose (FFFS) was used to discriminate pure minced beef meat and adulterated minced beef meat containing (1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100%) of chicken meat as an adulterant in uncooked beef meat samples. Fixed excitation (290 nm, 322 nm, and 340 nm) and fixed emission (410 nm) wavelengths were used for performing analysis. Fluorescence spectra were acquired from pure and adulterated meat samples to differentiate pure and binary mixtures of meat samples. Principle component analysis, partial least square regression and hierarchical cluster analysis were used as chemometric tools to find out the information from spectral data. These chemometric tools predict adulteration in minced beef meat up to 10% chicken meat but are not good in distinguishing adulteration level from 1% to 5%. The results of this research provide baseline for future work for generating spectral libraries using larger datasets for on-line detection of meat authenticity by using fluorescence spectroscopy.

• 대한기계학회논문집
• /
• 제17권9호
• /
• pp.2162-2171
• /
• 1993

Geneal Structure optimization is utilized to minimize the weight of structures while satisfying constraints imposed on stress, displacements and natural frequencies, etc. Sandwich structures consist of inside core and outside face sheets. The selected sandwich structures are isotropic sandwich beams and isotropic sandwich plate. The face sheets are treated as membrane and assumed to carry only tensions, while the core is assumed to carry only transverse shear. The characteristic of the varying area are considered by adding the projected component of the tension to the transverse shear. The bending theory and energy method are adopted for analyzing sandwich beams and plates, respectively. In the optimization process, the cost function is the weight of a structure, and a deflection and stress constraints are considered. Design variable are thickness and tapering coefficients which determine the shape of a structure. An existing optimization code is used for solving the formulated problems.

• 한국주거학회논문집
• /
• 제21권6호
• /
• pp.43-52
• /
• 2010

It has applied the design method of face to face dimension, an MC (Modular Coordination) design method, at design steps of an apartment house for standardization of architecture. But It has been still using center line size. A construction limit isn't specific on these center line size, and dead space occurs with excessive gaps as size matching isn't performed well between structural sieve and parts. And above all, measurement before construction isn't to expect standardization, and to increase custom-made furniture. We investigate manufacture of a preferential domestic system kitchen and construction state, and we derive from a problem in viewpoints of standardization for this. And we utilized a reference plane, and we present construction document and design plan in order to solve these problems. And we analyze an effect got by application of the assembly basis aspect that we presented, and we can raise efficiency of standardization of housing and construction industry, and we light ultimately up.

• International Journal of Control, Automation, and Systems
• /
• 제5권1호
• /
• pp.61-69
• /
• 2007

In this paper, we developed not only a reliable sound localization system including a VAD(Voice Activity Detection) component using three microphones but also a face tracking system using a vision camera. Moreover, we proposed a way to integrate three systems in the human-robot interaction to compensate errors in the localization of a speaker and to reject unnecessary speech or noise signals entering from undesired directions effectively. For the purpose of verifying our system's performances, we installed the proposed audio-visual system in a prototype robot, called IROBAA(Intelligent ROBot for Active Audition), and demonstrated how to integrate the audio-visual system.

• 한국정보통신학회논문지
• /
• 제6권2호
• /
• pp.346-355
• /
• 2002

이 연구는 인식의 정확성을 향상시키기 위하여 단일생체 인식 대신에 얼굴, 입술, 음성을 이용하는 다중생체 인식방법을 제안한다. 각 생체 특징은 다음과 같은 방법으로 찾는다. 얼굴 특징은 웨이블렛 다중분해와 주성분 분석방법으로 계산하였고, 입술의 경우는 입술의 경계를 구한후 최소 자승법을 이용한 방정식의 계수를 구하였으며, 음성은 멜 주파수에 의한 MFCC를 사용하였으며, 역전파 학습 알고리즘으로 분류하여 실험하였다. 실험을 통해 본 방법의 유효성을 확인하였다.

• 대한기계학회논문집
• /
• 제18권9호
• /
• pp.2211-2224
• /
• 1994

On face milling operation a new optimal cutter, which can minimize the resultant cutting forces, was designed from the cutting force model. Cutting experiments were carried out and the cutting forces of the new and conventional cutters were analyed in time and frequency domains. The resultant cutting forces were used as the objective function and cutter angles as the variables. A new optimal cutter design model which can minimize the resultant cutting forces under the constraints of variables was developed and its usefulness was proven. The cutting forces in feed direction of the newly designed cutter are reduced in comparison with those from the conventional cutter. The magnitudes of an insert frequency component of cutting force from the newly designed cutter are reduced than those from conventional cutter and the fluctuations of cutting force are also reduced.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 1998년도 춘계학술대회 논문집
• /
• pp.203-208
• /
• 1998

Materials of the Frame hole in fishing trackle reel is made up a number large and small holes. Thus, it is difficult to effective machining. Abrasive flow machining(AFM) is useful to polish a internal or external surface of the 3-dimensional shape parts, which are used in many fields such as aerospace, automative, semi-conductor, and medical component industries. The machining process is that two hydraulic cylinders, which are located face to face, enforce abrasive media to the passage between workpiece and tooling parts alternately, and then the abrasives include in the media pass the passage and polish the surface of workpiece. The media which is made of polymer and abrasives plays the role of the tool for deburring or polishing complex shap workpiece by its viscoelastic characteristics. In this study, the abrasive media for abrasive flow machining was made by mixing viscielastic polymer with alunina and silicon carbide abrasive. Also, we present AFM device design and preliminary results of an investigation in to some aspects of the AFM process performance in fishing trackle reel.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2011년도 제43차 동계학술발표논문집 19권1호
• /
• pp.315-316
• /
• 2011

일반적으로 영상에서 얼굴 영상을 검출하고 인식하는 알고리즘은 패턴 인식 연구에 있어서 인간과 컴퓨터의 상호작용의 연구라는 면에서 아주 중요한 문제로 연구되어 왔다. 본 논문에서는 고유얼굴을 이용하여 유클리디언 거리법과 퍼지기법의 인식률을 비교해보고자 한다. PCA(Principal Component Analysis) 방식은 우수한 인식 결과를 보장하는 얼굴인식 기법중의 하나이며, 얼굴 영상을 이용하여 공분산 행렬을 계산하고, 공분산 행렬을 통해 생성된 저차원의 벡터, 즉 고유얼굴(Eigenface)을 이용하여 가중치를 계산하고, 이 가중치를 기준으로 인식을 수행하는 기법이다. 이를 기반으로 하여, 본 논문에서는 전처리 과정, 고유얼굴 과정, 유클리디언 거리법 및 퍼지 소속도 함수 설계 과정, 신경망 학습과정, 인식과정으로 구성된 5단계의 얼굴 인식 알고리즘을 제안한다.