• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 컴퓨터소사이어티 추계학술대회논문집
• /
• pp.153-156
• /
• 2003

In this paper, we propose a new scheme for automatic segmentation of the liver in CT images. The proposed scheme is carried out on region of interest(ROI) blocks that include regions of the liver with high probabilities. The ROI approach saves unnecessary computational loss in finding the accurate boundary of the liver. The proposed method utilizes the composition of multi-size morphological filters with a prior knowledge, such as the general location or the approximate intensity of the liver to detect the initial boundary of the liver. Then, we make the gradient image with the weight of the initial liver boundary and segment the liver legion by using an immersion-based waters hed algorithm in the gradient image. finally, the refining process is carried out to acquire a more accurate liver region.

• 보존과학회지
• /
• 제29권1호
• /
• pp.69-79
• /
• 2013

납을 획득하는 방법과 처리과정에서 일어나는 물리 화학적 거동변화를 확인함으로써 납이 함유된 고대 금속유물의 제작기법 연구에 활용하고자 방연석을 현재 가행 중인 광산에서 채취하여 세척 및 선광 등의 전처리 공정을 거쳐 제련 및 정련실험을 실시하였다. 실험결과, 납은 배소-화학반응에 의해 추출되며, $1,000^{\circ}C$이상의 고온에서도 방연석이 matte 내 존재하고 있음을 알 수 있었다. 이후 잔존하고 있는 방연석을 제거하기 위해 정련실험을 실시한 결과, 약 $11.1g/cm^3$의 비중을 갖는 금속 납이 추출되었으며 방연석 내 미량으로 존재하고 있던 Ag가 입계 내에 응집되는 것을 확인하였다. 이를 통해 고순도의 납을 제작하기 위해서는 적어도 1번 이상의 정련 과정과 은을 제거하기 위한 회취법 등이 시행되었을 것이다.

• 한국멀티미디어학회논문지
• /
• 제7권11호
• /
• pp.1531-1539
• /
• 2004

본 논문은 입술의 경계선을 효과적으로 추출하는 방법을 제안하였다. 입술 형태는 PDM(Point Distribution Model)과 주성분 분석법을 이용하여 표현하고 입술 경계선은 GLDM(Gray Level Distribution Model)을 기반으로 표현하였다 입술 경계선 추출은 모델에 대한 입력영상의 정확도에 대한 목적함수를 최적화하는 문제로 단순화하였으며, 최적화를 위해 다운힐 심플렉스(Down Hill Simplex) 알고리즘을 이용하였다. 탐색과정에서 지역 최소점으로 수렴하는 문제를 해결하기 위하여 입술 형태 모델의 형태계수를 GMM(Gaussian Mixture Model)을 이용하여 표현하였다. 형태계수에 대한 GMM을 이용하여 입술의 대략적인 형태를 찾고, 이때 사용된 mixture 성분을 이용하여 탐색과정에서 입술의 형태를 조정함으로써 지역 최소점에 수렴하여 입술의 정확한 위치를 찾지 못하는 문제점을 해결하였다. 여러 영상을 대상으로 실험하여 좋은 결과를 얻었다.

• The Journal of the Acoustical Society of Korea
• /
• 제24권4E호
• /
• pp.128-131
• /
• 2005

A new algorithm for speech signal segmentation is proposed. This algorithm is based on finding successive similar frames belonging to a segment and represents it by an average spectrum. The speech signal is a slowly time varying signal in the sense that, when examined over a sufficiently short period of time (between 10 and 100 ms), its characteristics are fairly stationary. Generally this approach is based on finding these fairly stationary periods. Advantages of the. algorithm are accurate border decision of segments and simple computation. The automatic segmentations using frame averaging show as much as $82.20\%$ coincided with manually verified segmentation of CMU ARCTIC corpus within time range 16 ms. More than $90\%$ segment boundaries are coincided within a range of 32 ms. Also it can be combined with many types of automatic segmentations (HMM based, acoustic cues or feature based etc.).

• 한국주조공학회지
• /
• 제15권6호
• /
• pp.588-595
• /
• 1995

SiC particles reinforced Mg-Zr, Mg-Zn and Mg-Zn-Zr composites were manufactured by Rheocompocasting method. Effects of Zn, Zr addition on microstructures and hardness were investigated by using the micro Vickers hardness tester, the optical and scanning electron microscopy. By the Zr addition to the pureMg/SiCp composites, SiC particles become more homogeneously dispersed and grain refined so that the micro hardness of the composite increased. In case of Zn addition, although grain refinement and homogeneous dispersion effects of SiC particles were not obtained, hardness was more increased than the only Zr added composite by the formation of many Mg-Zn intermetallic compounds at grain boundary. In the Mg-Zn-Zr/SiCp composite, the highest value of hardness was obtained by triple effects such as grain refining, dispersion hardening of SiC particles and Mg-Zn compounds.

• 대한용접접합학회:학술대회논문집
• /
• 대한용접접합학회 2006년도 춘계 학술대회 개요집
• /
• pp.140-142
• /
• 2006

The bend ductility of Type 410S ferritic stainless steel overlay weld on carbon steel was investigated. Overlay weld that was stabilized with Nb had large columnar ferrite grain and Nb precipitate on grain boundary. And that caused fracture when bend test without concern of PWHT condition. Proper bend ductility at as-welded condition was achieved by refining ferrite grain with addition of $0.04{\sim}0.09%$ Al and $0.2{\sim}0.5%$ Ti that make oxide, carbide and nitride at high temperature.

• 전자공학회논문지D
• /
• 제34D권3호
• /
• pp.74-86
• /
• 1997

With continued miniaturization and development of new devices, the highly nonuniform oxidation of three-dimensional non-planar silicon structures plays an increasingly important role. In this paper, the three-dimensional finite element numerical simulator. Grwoth of oxide is a coupled process of diffusion of oxidant and deformation of oxide. Because boundaries of oxide are moved in each time step and LOCOS structure is formed three-dimensional shape of sruface, it is necessary to develope an efficient node control algorithm that can locally generate and eliminate the node. Therefore we have developed the optimized three-dimensional mesh generator which is cpable of refining and eliminating the meshes at the moving boundary of oxide, and hve developed three-dimensional finite element oxidation solver.

• International journal of advanced smart convergence
• /
• 제12권4호
• /
• pp.134-141
• /
• 2023

We present a novel method aimed at refining ground truth data through regularization and modification, particularly applicable when working with the original ground truth set. Enhancing the performance of deep neural networks is achieved by applying regularization techniques to the existing ground truth data. In many machine learning tasks requiring pixel-level segmentation sets, accurately delineating objects is vital. However, it proves challenging for thin and elongated objects such as blood vessels in X-ray coronary angiography, often resulting in inconsistent generation of ground truth data. This method involves an analysis of the quality of training set pairs - comprising images and ground truth data - to automatically regulate and modify the boundaries of ground truth segmentation. Employing the active contour model and a recursive ground truth generation approach results in stable and precisely defined boundary contours. Following the regularization and adjustment of the ground truth set, there is a substantial improvement in the performance of deep neural networks.

• 디지털융복합연구
• /
• 제13권11호
• /
• pp.481-491
• /
• 2015

본 연구는 대학의 인성교육을 목표로 개발된 교양교과목의 효과성을 검증하고 결과를 논의함으로써, 향후 개선된 인성교육 교과를 개발하기 위함에 목적이 있다. 연구는 지방의 K대학교의 마음나눔교육부에서 대학의 교육 목표에 따라 개발된 '마음교양' 교양교과를 대상으로 효과성을 검증하였다. 교과는 15주로 진행했으며 구체적인 인성교육의 내용은 '감사'의 가치관 증진으로 구성되었다. 효과성을 검증하기 위해 켄터키 마음챙김기술 척도, 게슈탈트접촉경계진단검사, 나의식우리의식 척도, 자존감 척도를 사용하였다. 연구대상자는 '마음교양' 교양교과를 수강한 학생(실험군)과 수강하지 않은 학생(통제군) 229명의 자료이며, 연구방법은 집단간 차이분석을 하였다. 분석결과, 게슈탈트접촉경계진단검사의 하위요인인 연결성 척도에서만 유의미한 집단간 차이를 보였다(t=2.323, p=.021). 이는 대인관계에서의 연결성 회복을 시사하며, 인성교육의 일부 효과성이 검증되었음을 의미한다. 추후 연구에서는 감사의 마음이외에 K대학교의 교육이념을 반영한 교과 내용을 구체화하고, 교육방법에서도 K여자대학의 교육 목표를 구체적으로 반영한 교과개발이 필요함이 논의되었다.

• 한국주조공학회지
• /
• 제22권6호
• /
• pp.281-287
• /
• 2002

In order to investigate the variation of microstructure and property of the Electro-slag Remelted M2 steel, microstructure observation, hardness, and bending test were performed by using optical microscope. SEM/EDS, rockwell hardness tester, charpy impact tester and bending tester, respectively. It was revealed that the number of inclusions and content of gas elements(S, O, N) in M2 steel fabricated by ESR process decreased markedly compared to those of AIM. It seems to be due to refining effect of ESR process. The volume fraction of carbides in quenched and tempered specimens after austenitizing at 1150$^{\circ}C$ and 1240$^{\circ}C$ was measured. The volume fraction of grain boundary carbides were found to be similar for both specimens. However, The volume fraction of carbides in grain decreased with an increase of austenitizing temperature. When specimen was austenitized at 1150$^{\circ}C$, grain boundary carbides showed needle like morphology. But, the carbides were broken with an increase of austenitizing temperature. The specimen austenitized at 1240$^{\circ}C$ showed higher hardness and lower bending strength compared to that of 1150$^{\circ}C$. As expected, toughness increased with sub-zero quenching treatment.