• 한국지능시스템학회논문지
• /
• 제22권6호
• /
• pp.748-753
• /
• 2012

본 논문에서는 최적 pRBFNNs 패턴분류기 기반 3차원 스캐너를 이용한 얼굴인식 알고리즘을 설계한다. 일반적으로 2차원 영상을 이용한 얼굴인식 시스템은 사진의 명암도를 이용하여 얼굴의 특징을 추출하게 된다. 그렇기 때문에 빛이나 조명, 또는 얼굴 포즈와 같은 환경 변화들은 시스템의 성능을 저하시킨다. 따라서 본 논문에서 제안된 얼굴인식 알고리즘은 2차원 얼굴인식 시스템의 한계를 극복하기 위하여 3차원 스캐너를 사용하여 설계한다. 먼저 3차원 스캐너를 이용하여 얼굴 형상을 스캔하고 스캔된 얼굴 형상은 포즈 보상 과정을 통하여 정면으로 변환된다. 그 후에 Point Signature 기법을 사용하여 얼굴의 깊이 정보를 추출하고 마지막으로 고차원 패턴인식 문제에 대한 해결을 위하여 최적화된 pRBFNNs (Polynomial-based Radial Basis Function Neural Networks) 모델을 사용하여 인식성능을 확인한다.

• 한국정밀공학회지
• /
• 제22권9호
• /
• pp.188-193
• /
• 2005

To provide the various machining materials with excellent quality and dimensional accuracy, high -speed machining is very useful tool as one of the most effective rapid manufacturing processes. However, high-speed machining is not suitable for microscale thin-walled structures because of the lack of the structure stiffness to resist the cutting force. A new method which is able to make a very thin-walled structure rapidly will be proposed in this paper. This method is composed two processes, high-speed machining and filling process. Strong workholding force comes out of the solidification of filling materials. Low-melting point metal alloys are used in order to minimize the thermal effect during phase change and to hold arbitrary shape thin-walled structures quickly during high-speed machining. To verify the usefulness of this method, we will show some applications, for examples thin -wall cylinders and hemispherical shells, and compare the experimental results to analyze the dimensional accuracy of typical parts of the structures.

• E2M - 전기 전자와 첨단 소재
• /
• 제7권2호
• /
• pp.107-116
• /
• 1994

In the manufacturing of VLSI circuits, variations of device characteristics due to the slight differences in process parameters drastically aggravate the performances of fabricated devices. Therefore, it is very important to establish optimal process conditions in order to minimize deviations of device characteristics. In this paper, we used one-dimensional process simulator, SUPREM-II, and two dimensional device simulator, MINIMOS 4.0 in order to extract optimal process parameter which can minimize changes of the device characteristics caused by process parameter variation in the case of short channel nMOSFET and pMOSFET device. From this simulation, we have derived the dependence relations between process parameters and device characteristics. Here, we have suggested a method to extract process parameters from design trend curve(DTC) obtained by these dependence relations. And we have discussed short channel effects and device limitations by scaling down MOSFET dimensions.

• 대한수학회논문집
• /
• 제38권2호
• /
• pp.561-573
• /
• 2023

In this paper, we extend the medial triangle theorem and Varignon's theorem to generic two-dimensional polygons and highlight the role played by diagonals in this process. One of the results is a synthetic definition of the concept of median for an n-sided polygon.

• 한국광학회지
• /
• 제14권5호
• /
• pp.504-508
• /
• 2003

최근 전자산업계에 새롭게 널리 생산되는 마이크로 전자부품들은 왜곡이 최소화된 정밀한 외관 형상을 갖도록 제조되고 관리되지만, 측정 대상의 표면이 가시광 영역에서 광산란되는 특징을 가짐으로 인해, 기존의 피죠나 마이켈슨 형태의 비교간섭법으로는 고정밀의 삼차원 형상측정이 용이하지 아니하였다. 본 논문에서는 광섬유를 이용한 새로운 개념의 점회절 간섭계를 제안하고, 이를 광산란 거친표면의 대표적인 제품인 칩패키지와 실리콘 웨이퍼의 삼차원 형상 측정에 적용하였다. 측정결과 66 mm 측정영역에서 측정 형상오차 PV(peak-to-valley value) 5.6 $\mu\textrm{m}$, 분산값($\sigma$) 1.5 $\mu\textrm{m}$를 획득함으로써 기존의 비교 간섭 측정법에 비해 더욱 향상된 측정 정밀도를 획득하였다.

• 제어로봇시스템학회논문지
• /
• 제21권10호
• /
• pp.965-971
• /
• 2015

This paper suggests a feature point-based Iterative Closest Point (ICP) algorithm to compensate for the disparity error in building a two-dimensional map. The ICP algorithm is a typical algorithm for matching a common object in two different images. In the process of building a two-dimensional map using the laser scanner data, warping and distortions exist in the map because of the disparity between the two sensor values. The ICP algorithm has been utilized to reduce the disparity error in matching the scanned line data. For this matching process in the conventional ICP algorithm, pre-known reference data are required. Since the proposed algorithm extracts characteristic points from laser-scanned data, reference data are not required for the matching. The laser scanner starts from the right side of the mobile robot and ends at the left side, which causes disparity in the scanned line data. By finding the matching points between two consecutive frame images, the motion vector of the mobile robot can be obtained. Therefore, the disparity error can be minimized by compensating for the motion vector caused by the mobile robot motion. The validity of the proposed algorithm has been verified by comparing the proposed algorithm in terms of map-building accuracy to conventional ICP algorithm real experiments.

• 대한기계학회논문집B
• /
• 제40권7호
• /
• pp.471-475
• /
• 2016

본 연구는 의료용 동위원소 생산공정에서 발생하는 방사성폐기물을 저장하는 장기저장소의 온도를 두 가지 방법으로 평가한 결과를 보여준다. 방사성폐기물에서 발생하는 열을 Volume source와 Point source로 가정하여 장기저장소의 온도를 평가한 결과, 폐기물 저장위치에 따른 최대온도분포는 유사하게 나타났으나 그 크기에 있어서 최대 $5^{\circ}C$ 정도의 차이를 보였다. 따라서, 개념설계를 위해서는 해석 시간이 오래 걸리는 Volume source를 이용한 3차원 해석보다는 Point source를 이용한 2차원 해석이 보다 효율적이지만, 상세 설계를 위한 정확한 해석 결과를 얻기 위해서는 Volume source를 이용한 3차원 해석이 수반되어야 함을 알 수 있다.

• 제어로봇시스템학회논문지
• /
• 제20권12호
• /
• pp.1267-1271
• /
• 2014

There have been many autonomous robot calibration methods which form closed loop structures through the various attached sensors and mechanical fixtures. Single point calibration among them has been used for on-site calibration due to its convenience of implementation. The robot can reach a single point with infinitely many configurations so that single point calibration algorithm can be set up and easily implemented relative to the other methods. However, it is not still easy to drive the robots' sharp edge to its corresponding edge of the fixture. This is error-prone process. In this paper, we propose a 3 dimensional small displacement measuring sensor and a robot calibration algorithm based on this sensor. This method relieves the difficulty of matching two edges in the single point calibration and improves the resulting robot accuracy. Simulated study is carried out on a Hyundai HA06 robot to show the effectiveness of the proposed method over the single point calibration. And also, the resulting robot accuracy is compared with that from 3D laser tracker based calibration to show the dependency of robot accuracy on range of the workspace where the measurement data are collected.

• 디자인학연구
• /
• 제13권
• /
• pp.105-113
• /
• 1996

The great development of the modern science enables the information technology including computer to be promoted, which has the great effects on the whole design field. Most of all, the computer has the feature to express the mathmatical formular through the logic and system, which is being used very efficiently in the design work. CG application to the two-dimensional design may be limited in promoting the basic ability synthetically through the direct and molding experience for its attribute, however, it is possible to create the effects on the different level in addition to those from the traditional skill. In particular, we should keep the fact in mind that the CG application in the design field will be able to accomplish the object only under understanding the attributes and functions for it to have. We should carry out the task to agree with. First, it is necessary to develop the various study on the educational environment to make the process to solve the problems and to promote the effects of the educational media in teaching design with the CG application. This study would make plans to approach that the computer can be utilized for the main method to create design despite of being limited in some point. Moreover, it suggests the possibility for the CG application in the course of building up and motiving the models by applied the basic system, which enables the computer to be applied centering around the composition of geometry pattern of the basic course of the two-dimensional design.

• 대한기계학회논문집B
• /
• 제23권8호
• /
• pp.997-1009
• /
• 1999

Silica particle formation and growth process including chemical reaction, coagulation and sintering was studied in a counterflow diffusion flame burner. The counterflow geometry provides a one dimensional flow field, along the stagnation point streamline, which greatly simplifies interpretation of the particle growth characteristics. $SiCl_4$ has been used as the source of silicon in hydrogen/oxygen/argon flames. The temperature profiles obtained by calculation showed a good agreement with experiment data. Using one and two dimensional sectional method, aerosol dynamics equation in a flame was solved, and these two results were compared. The two dimensional section method can consider sintering effect and growth of primary particle during synthesis, thus it showed evolution of morphology of non-spherical particles (aggregates) using surface fractal dimension. The effects of flame temperature and chemical loading on particle dynamics were studied. Geometric mean diameter based on surface area and total number concentration followed the trend of experiment results, especially, the change of diameters showed the sintering effect in high temperature environment.