• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2003년도 봄 학술발표회 논문집
• /
• pp.288-295
• /
• 2003

An adaptive procedure in finite element analysis is presented by p-refinement of meshes in conjunction with a posteriori error estimator that is based on the recovery technique. In case of the recovery technique, the SPR(superconvergent patch recovery) approach has been modified for p-adaptive mesh refinement. The strategy of finding a nearly optimal distribution of polynomial degrees on a fixed finite element mesh is discussed such that a particular element has to be refined automatically to obtain an acceptable level of accuracy by increasing p-levels non-uniformly. To verify the proposed algorithm, the limit value approach is proposed which utilizes the exact strain energy computed from the extrapolation equation. A new pre-processor is developed for the p-version finite element program in which the vector graphic editor is used for the automatic generation of node connection and coordinate by halfedge solid data structure according to uniform or nonuniform p-distribution. The general 2-D algorithm is also developed to generate face modes and internal modes in accordance with different mesh types. The quality of the error estimator is investigated with the help of two mumerical examples. The results show that the sequences of p-distributions obtained by the proposed error indicator closely follow the optimal trajectory.

• Korean Journal of Geomatics
• /
• 제2권1호
• /
• pp.83-89
• /
• 2002

The cadastral maps as many as about 750,000 map sheets to cover 34,751,000 parcels of land of Korea, are being digitalized. The problem of shrinkage-expansion of the paper cadastral maps has to be resolved for the new digital maps, where the nodes and vertices of the parcel boundaries are represented by coordinates. The photo coordinate refinement techniques, two dimensional projective transformation and local area transformation as in the reseau grid method, were introduced for this distortion correction. Using the fact that original maps drawn on the plane tables in field from 1910 to 1918 have grid lines and have been preserved well, a strategic flow to apply the refinement techniques to the digital maps with the original maps as controls was developed. To accommodate the presence or absence of the original maps and grid lines, and different scales and sizes of the maps, the strategy was implemented by a computer program package. Various distortions and corrections were simulated and errors were evaluated. The RMS errors in the corrected digital maps were allowable, thus, the method developed in this study was to be applicable for the digital cadastral maps.

• 한국컴퓨터정보학회논문지
• /
• 제27권12호
• /
• pp.77-84
• /
• 2022

본 논문은 객체 위치식별 알고리즘의 성능을 향상하기 위한 레이블 재할당 방법을 제안한다. 제안한 방법은 추론 단계와 재할당 단계로 구분한다. 추론 단계에서는 학습된 모델로부터 사전 지정된 크기에 따라 다중 스케일 추론을 수행한 뒤, 이를 마스킹한 영상을 다시 한번 추론하여 강인한 클래스 종류의 추론 결과를 얻는다. 재할당 단계에서는 박스간의 IoU를 계산하여 중복 박스를 제거하고, 박스와 클래스의 빈도를 계산하여 지배적 클래스를 다시 할당하였다. 제안한 방법을 검증하기 위하여 공사현장 안전장비 인식 영상 데이터 세트에 레이블 재할당 방법을 적용하고 이를 YOLOX-L 객체 탐지 모델에서 학습하였다. 실험 결과 적용 전 대비 mAP가 3.9% 향상하여 51.07%를 달성하였으며 AP_S를 3배 이상 향상하여 14.53%를 달성하였다. 실험 결과를 통해 레이블 재할당 알고리즘이 더 우수한 성능의 모델을 훈련해 냄을 확인하였다.

• Structural Engineering and Mechanics
• /
• 제62권5호
• /
• pp.537-550
• /
• 2017

Structural sizing is a rewarding task due to its non-convex constrained nature in the design space. In order to provide both global exploration and proper search refinement, a hybrid method is developed here based on outstanding features of Evolutionary Computing and Teaching-Learning-Based Optimization. The new method introduces an observer phase for memory exploitation in addition to vector-sum movements in the original teacher and learner phases. Proper integer coding is suited and applied for structural size optimization together with a fly-to-boundary technique and an elitism strategy. Performance of the proposed method is further evaluated treating a number of truss examples compared with teaching-learning-based optimization. The results show enhanced capability of the method in efficient and stable convergence toward the optimum and effective capturing of high quality solutions in discrete structural sizing problems.

• International Journal of Aeronautical and Space Sciences
• /
• 제8권1호
• /
• pp.66-78
• /
• 2007

A multi-point aerodynamic shape optimization technique has been developed for helicopter rotor blades in hover based on a continuous adjoint method on unstructured meshes. The Euler flow solver and the continuous adjoint sensitivity analysis were formulated on the rotating frame of reference. The 'objective function and the sensitivity were obtained as a weighted sum of the values at each design point. The blade section contour was modified by using the Hicks-Henne shape functions. The mesh movement due to the blade geometry change was achieved by using a spring analogy. In order to handle the repeated evaluation of the design cycle efficiently, the flow and adjoint solvers were parallelized based on a domain decomposition strategy. A solution-adaptive mesh refinement technique was adopted for the accurate capturing of the wake. Applications were made to the aerodynamic shape optimization of the Caradonna-Tung rotor blades and the UH-60 rotor blades in hover.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2006년도 학술대회 2부
• /
• pp.789-794
• /
• 2006

UI (User Interface)는 끊임없이 진화한다. 사용자의 새로운 요구 사항을 만족시키기 위해, 기존 UI에 대한 불편 사항을 해결하기 위해, 경쟁 제품의 더 좋은 UI를 따라 잡기 위해, 그리고 새로운 기술로 기존에는 불가능했던 것이 가능해 짐으로 인해, UI는 지속적으로 더 좋은 방향으로 변화하고 있다. 특히 휴대폰과 같이 라이프사이클이 짧은 제품의 UI는 매 제품마다 UI를 변화시킬 수 없기 때문에 기업 입장에서의 전략적인 접근이 중요해 진다. 더욱이 충분한 검증 없이 이루어진 진화의 방향은 사용자가 원하는 방향과 일치하지 않게 되어 오히려 퇴화하는 문제도 발생할 수 있다. 본 논문은 UI가 생성, 유지, 융합, 소멸되어 가는 진화론적 발전 속성 및 그러한 속성에 영향을 주는 요소들을 살펴보고, 라이프사이클이 짧은 제품의 UI를 개선할 때 어떤 전략을 취하는 것이 기업 차원에서 좋은지를 설명한다. UI 진화 전략은 크게 점진적 개선을 추구하는 구릉식 전략과 한번에 도약하는 계단식 전략으로 나눌 수 있다. 위 두 전략들은 사용자 만족도, 개선에 필요한 비용과 그 가치, 개선 후의 리스크 (Risk)에 따라 장단점을 가진다. 결론적으로 본 논문에서는 두 전략 모두 서로 상반되는 장단점을 가지므로, 어느 한 전략을 취하기 보다는 UI 요소의 특성 및 변화의 규모에 따라 두 전략을 혼용하는 전략이 기업 입장에서 최적의 전략임을 주장한다.

• 정보기술과데이타베이스저널
• /
• 제4권2호
• /
• pp.33-45
• /
• 1998

Database summarization is the procedure to obtain generalized and representative descriptions expressing the content of a large amount of database at a glance. We present a top-down summary refinement procedure to discover database summaries. The procedure exploits attribute concept hierarchies that represent ISA relationships among domain concepts. It begins with the most generalized summary and proceeds to find more specialized ones by stepwise refinements. This top-down paradigm reveals at least two important advantages compared to the previous bottom-up methods. Firstly, it provides a natural way of reflecting the user's own discovery preference interactively. Secondly, it does not produce too large intermediate result that makes it hard for the bottom-up approach to be applied in practical environment. The proposed procedure can also be easily extended for distributed databases. Information content measure of a database summary is derived in order to identify more informative summaries among the discovered results.

• 패션비즈니스
• /
• 제23권5호
• /
• pp.81-95
• /
• 2019

The purpose of this study is to research consumers' perception of the HiSeoul fashion show, which is being used by new designers as a means of promotion, and to propose a strategy for revitalizing new designer brands. This was done in order to secure basic data from fashion consumers, to help guide marketing strategies and promote rising designers. In this research, the consumers' perception of HiSeoul fashion show was verified using text-mining, data refinement and word clouding that was undertaken by TEXTOM3.0. Also, semantic network analysis, CONCOR analysis and visualization of the analysis results were performed using Ucinet 6.0 and NetDraw. "HiSeoul fashion show" was used as the keyword for text-mining and data was collected from March 1, 2018 to April 30, 2019. Using frequency analysis, TF-IDF, and N-gram, it was also shown that consumers are aware of places where shows are held, such as DDP and Igansumun. It was also revealed that consumers recognize rising designer brands, designer's names, the names of guests attending the show and the photo times. This study is meaningful in that it not only confirmed consumers' interest in new designer brands participating in the HiSeoul Fashion Show through big data but also confirmed that it is available as a marketing strategy to boost brand sales. This study suggests using HiSeoul show room to induce consumer sales, or inviting guests that match the brand image to promote them on SNS on the day the show is held for a marketing strategy.

• Computers and Concrete
• /
• 제27권3호
• /
• pp.283-295
• /
• 2021

Existing meso-scale models of concrete need to refine the mesh grids of aggregate and cement mortar, which may greatly reduce the computational efficiency. To overcome this problem, a novel meso-scale modeling strategy, which is based on rigid body spring method and Voronoi diagram, is proposed in this study to establish the meso-scale model of concrete. Firstly, establish numerical aggregate models according to user-defined programs. Circle aggregates are adopted due to their high efficiency in generation and packing process, and the grading of aggregate are determined according to the distribution curve proposed by Full and Thompson; Secondly, extract the centroids of aggregates, and then develop the Voronoi diagram in which aggregate centroids are defined as initial scatters; Finally, establish the rigid body spring model for concrete based on the Voronoi diagram. Aggregates are represented by rigid blocks, and assumed to be unbreakable. Cement mortar is concentrated into the interface between adjacent blocks and represented by two uniform springs. The number of grids is consistent with that of aggregates in specimens, and no mesh-refinement of aggregates and cement mortar is required. The accuracy and efficiency of the proposed modeling strategy are firstly identified by comparing the numerical results with the experimental ones, and then the applicability of the proposed strategy with different volume percentage occupied by aggregates is investigated.

• 한국콘텐츠학회논문지
• /
• 제17권3호
• /
• pp.329-336
• /
• 2017

비디오 압축에서 움직임 추정 기법은 영상 화질과 발생 비트량에 있어서 중요한 역할을 하지만 많은 계산 복잡도를 요구한다. 다수의 카메라로 촬영한 동영상인 다시점 비디오는 카메라의 수에 비례하여 데이터의 양이 기하급수적으로 증가하기 때문에 움직임 추정에 많은 계산량을 필요하다. 본 논문에서 다시점 비디오의 부호화를 위한 움직임 추정의 계산량을 줄이면서 화질과 비트량을 유지하는 움직임 추정 기법을 제안한다. 제안한 움직임 추정 기법은 움직임 벡터들의 분포 특성과 영상 블록들의 움직임 특성을 이용한다. 제안한 움직임 추정기법은 계층적 움직임 추정 기법으로 대칭형 멀티 마름모 패턴, 대각선 패턴, 사각형 패턴 그리고 정교한 패턴으로 구성되어 있다. 제안한 움직임 추정 기법은 움직임 벡터들과 블록 움직임의 특성들을 이용하여 패턴들의 탐색 점들을 탐색 영역 내에 대칭적으로 배치하고 블록 움직임 크기에 따라 적응적으로 탐색 패턴을 선택하여 움직임 벡터를 추정한다. 제안한 기법의 성능은 JMVC의 고속 움직임 추정 기법인 TZ 탐색 기법과 전역 탐색 기법인 PBS (Pel Block Search)의 성능과 비교한 경우, 영상 화질면에서와 발생 비트량면에서 비슷하지만 움직임 추정에 필요한 계산량을 각각 약 40~75%, 98%감소시킨다.