• Computational Structural Engineering
• /
• v.8 no.3
• /
• pp.103-111
• /
• 1995

Evaluating nonlinear response of a MDOF system under dynamic stochastic loads such as seismic excitation usually requires excessive computational efforts. To alleviate this computational difficulty, an approximation is developed in which the MDOF inelastic system is replaced by a simple nonlinear equivalent system(ENS).Me ENS retains the most important properties of the original system such as dynamic characteristics of the first two modes and the global yielding behavior of the MDOF system. The system response is described by the maximum global(building) and local(interstory) drifts. The equivalency is achieved by two response scaling factors, a global response scaling factor R/sub G/, and a local response scaling factor R/sub L/, applied to the responses of the ENS to match those of the original MDOF system. These response scaling factors are obtained as functions of ductility and mass participation factors of the first two modes of structures by extensive regression analyses based on results of responses of the MDOF system and the ENS to actual ground accelerations recorded in past earthquakes. To develop the ENS with two response scaling factors, Special Moment Resisting Steel Frames are considered. Then, these response scaling factors are applied to the response of ENS to obtain the nonlinear response of MDOF system.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.9
• /
• pp.75-84
• /
• 2012

HMM-based Text-to-Speech systems generally utilize context dependent tri-phone units from a large corpus speech DB to enhance the synthetic speech. To downsize a large corpus speech DB, acoustically similar tri-phone units are clustered based on the decision tree using context dependent information. Context dependent information includes phoneme sequence as well as prosodic information because the naturalness of synthetic speech highly depends on the prosody such as pause, intonation pattern, and segmental duration. However, if the prosodic information was complicated, many context dependent phonemes would have no examples in the training data, and clustering would provide a smoothed feature which will generate unnatural synthetic speech. In this paper, instead of complicate prosodic information we propose a simple three prosodic boundary types and decision tree questions that use rising tone, falling tone, and monotonic tone to improve naturalness. Experimental results show that our proposed method can improve naturalness of a HMM-based Korean TTS and get high MOS in the perception test.

• The Journal of the Korea Contents Association
• /
• v.20 no.1
• /
• pp.142-151
• /
• 2020

The reason why today's public actively uses social communication is that the necessary information is collected and classified under the name of social big data through the web space to create the big data era, an ecosystem of information. In order for big data information to be used by the public, it is necessary to visualize it easily. This study categorized the types of visualization according to the information of social big data, and targeted the experienced students including the related majors and the general public who need to directly utilize and study the actual big data visualization as an experience evaluation target. As a result of analyzing the experiences of the experienced people, important implications for the visualization method for managing, analyzing, and utilizing the data were derived. The big data visualization strategy is to be expressed in a way that fits the data environment and user's eye level on SNS. In the future, if big data visualization is applied to product service or social trend, it will be an important data in terms of broadening its role, scope of application, and application.

• Journal of Biomedical Engineering Research
• /
• v.25 no.1
• /
• pp.43-47
• /
• 2004

Interests in human brain functionality and its connectivity have much frown up. DTI (Diffusion tensor imaging) has been known as a non-invasive MR) technique capable of providing information on water diffusion in tissues and the organization of white matter tract. Thus. It can provide us the information on the direction of brain fiber tract and the connectivity among many important cortical regions which can not be examined by other anatomical or functional MRI techniques. In this study. was used the 24 bit color coding scheme on the IDL platform in the windows environment to visualize the orientation of major fiber tracts of brain such as main association, projection, commissural fibers and corticospinal tracts. We additionally implemented a color coding scheme for each directional component and FA (fractional anisotropy), and used various color tables for them to be visualized more definitely. Consequently we implemented a fancy and basic technique to visualize the directional information of fiber tracts efficiently and we confirmed the feasibility of the 24 bit color coding scheme in DTI by visualizing main fiber tracts.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.33 no.3
• /
• pp.66-71
• /
• 2015

Given the recent trend of natural planting, the recognized needs for new landscaping plants that have advantages in terms of climate change and maintenance, and expected increases in demands for grasses in Korea, this study is intended to investigate from the design point of view the techniques to use grasses and their significance through garden design by Piet Oudolf who is attracting international interests with the use of perennial plants and grasses and is leading the trends in modern planting design, thereby answering the question: how to best use grasses in landscaping spaces? The characteristics of Oudolf's garden design using grasses are summarized in the following conclusions: First, Oudolf combines perennial plants and grasses to make one-to-one correspondences or express expanded drifts. Here grasses mainly serve as an element to change over to other spaces or as a connecting element between image transitions. Second, the brown color and texture of grasses represent Oudolf's considerations on the temporal continuity of gardens. They express the lyricism and pictorialism of autumn and winter. Third, grasses serve to set layers in wide areas resulting in discordance between viewpoints and circulations. Oudolf repeatedly cross perennial plants and grasses using matrices, islands and distributed layering. Here grasses are used to express abstractive meanings in the settings of scenes.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.30 no.2
• /
• pp.46-63
• /
• 2012

The object of this study was to analyze the speciality of Korean traditional waterscape and unique landscape formed with reflection of the phases of the time sat the area of Juhabru(宙合樓) in Changdeok Palace as a basic research to find the prototype of Taeaek pond at Changdeok Palace and restoration of the palace. Originally, Taeaek pond at Changdeok Palace was constructed in KingInjo(仁祖) period as a name of Yongji(龍池), later it called Taeaek pond after King Sukjong(肅宗). There is an island as a symbol of the immortal isle, and Chungseojeong(淸署亭), Taiksujae(澤水齋) and Buyongjeong(芙蓉亭) which were built to view the waterscape in Taeaek pond. Buildings were built asymmetrical balance around Taeaek pond because of the morphological character of tetragonal pond. Arrangement of this area has a definite form of axial structure. Yeolgokwan(閱古觀) Gaeyuwa(皆有窩), pavilions, bridges, islands, Osumoon(魚水門) and Juhabru are located on the north and south axis, and island and Osumoon play a role as a intersection and form an east of west axis. In this study, manual of construction for an island and pavilions is provided by analyzing transformational process of island and pavilions at Taeaek pond. Furthermore, kings and officials used to statically enjoy the view around Taeaek pond area, but dynamic fishing and boating activity happened in King Jungjo(正祖) period. These historical backgrounds have an influence on the spatial organization of Taeaek pond. For instance, bridge between Taeaek pond and island was destroyed with the increase of the importance of boating. Symbolic structure around Taeaek pond means 'fish changes to dragon' and 'both of king and officials become one'. Taiksujai, carving fish, Osumoon and Juhabru are provided as a related spatial factors.

• Geophysics and Geophysical Exploration
• /
• v.12 no.4
• /
• pp.361-366
• /
• 2009

Layered-earth Green's functions in electormagnetic (EM) surveys play a key role in modeling the response of exploration targets. They are computed through the Hankel transforms of analytic kernels. Computational precision depends upon the choice of algebraically equivalent forms by which these kemels are expressed. Since three-dimensional (3D) modeling can require a huge number of Green's function evaluations, total computational time can be influenced by computational time for the Hankel transform evaluations. Linear digital filters have proven to be a fast and accurate method of computing these Hankel transforms. In EM modeling for 3D inversion, electric fields are generally evaluated by the secondary field formulation to avoid the singularity problem. In this study, three components of electric fields for five different sources on the surface of homogeneous half-space were derived as primary field solutions. Moreover, reflection coefficients in TE and TM modes were produced to calculate EM responses accurately for a two-layered model having a sea layer. Accurate primary fields should substantially improve accuracy and decrease computation times for Green's function-based problems like MT problems and marine EM surveys.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2004.03a
• /
• pp.435-440
• /
• 2004

연안침수지도는 태풍이나 해일에 의한 연안침수 발생 시 방조제의 월류 및 붕괴로 인한 예상 침수구역을 강우빈도(200년)별로 나타내고, 침수면적과 깊이를 표현한 지도로서 국외 선진국에서 방재형 국토관리 정책결정과 침수피해에 대한 대민 홍보의 수단으로 활용되고 있다. 국내에서도 연안침수에 의한 피해가 나날이 커지고 국토의 개발에 따른 자연 재해가 많아짐에 따라 보다 방재형 국토개발로 국민의 안전과 복리를 증진하고, 연안침수에 의한 피해를 최소화하기 위한 노력의 일환으로 연안침수지도의 제작 필요성이 증대되고 있다. 본 연구의 시범지역인 대부도 남부지역의 GIS 데이터베이스 구축은 문헌조사 및 기구축 데이터 확보, 지역의 특성에 맞는 신규 GIS 데이터베이스 구축 방법을 제시하였다. 침수구역을 예측하기 위한 모델링을 지원하기 위한 정확도와 경제성, 객관성, 현실성이 높은 지형공간자료의 효율적인 구축이 강조되었다. 구축한 데이터베이스의 세부 유형은 침수위험구역, 침수경계구역, 침수량 화살표이며, 기존에 구축된 데이터를 활용한 레이어는 해안선, 도로, 건물, 등고선, 등심선, 연안 표고점 측량데이터이다. 연안침수지도는 효율적인 재난관리를 위하여 다양하게 활용될 수 있을 것으로 사료되며 나아가 주민의 침수에 대한 경각심 고취와 사전 교육, 지방자치단체의 침수위험 관리업무의 활용이 가능할 것으로 사료된다. 본 연구에서 제작된 연안침수지도는 지자체 주민들을 위해 보기 쉽고 다양한 정보를 얻을 수 있도록 제작되었으며, 이를 위해 사진과 삽화를 추가하여 주민들에게 침수피해에 대한 경각심과 이해를 높이는데 중점을 두었다. 유지관리 후 활용체계 정비의 중요성’을 점검하고, ‘변동자료 발생원인의 분석’, ‘명확한 변동자료 입력 근거의 확보’, ‘갱신주체별 역할의 정의 및 유지관리 기준의 설정’, ‘분야별업무 특성을 고려한 관련 기준의 마련 및 타 시스템과 연계되는 항목을 고려한 절차 정의’ 등에 대한 다양한 접근을 시도하였다. 본 연구에서 제시하는 유지관리 모델을 기반으로 각 지자체별로 적절한 컨설팅이 진행되고 이에 따라 담당자의 실천이 이루어진다면 지자체 GIS의 투자대비 효과에 대한 기대는 이상이 아닌 현실로 다가오게 될 것이다.가오게 될 것이다. 동일하게 25%의 소유권을 가지고 있다. ？스굴 시추사업은 2008년까지 수행될 계획이며, 시추작업은 2005년까지 완료될 계획이다. 연구 진행과 관련하여, 공동연구의 명분을 높이고 분석의 효율성을 높이기 위해서 시료채취 및 기초자료 획득은 4개국의 연구원이 모여 공동으로 수행한 후의 결과물을 서로 공유하고, 자세한 전문분야 연구는 각 국의 대표기관이 독립적으로 수행하는 방식을 택하였다 ？스굴에 대한 제1차 시추작업은 2004년 3월 말에 실시하였다. 시추작업 결과, 약 80m의 시추 코아가 성공적으로 회수되어 현재 러시아 이르쿠츠크 지구화학연구소에 보관중이다. 이 시추코아는 2004년 8월 중순경에 4개국 연구팀원들에 의해 공동으로 기재된 후에 분할될 계획이다. 분할된 시료는 국내로 운반되어 다양한 전문분야별 연구에 이용될 것이다. 한편, 제2차 시추작업은 2004년 12월에서 2005년 2월 사이에 실시될 계획이다. 수백만년에 이르는 장기간에 걸쳐 지구환경변화 기록이 보존되어 있는

• Spatial Information Research
• /
• v.21 no.6
• /
• pp.23-32
• /
• 2013

Recently, there has been much interest in the converged platform that enables the harmonized collection, integration, exchange, presentation and analysis of various kinds of marine information by using the ICT means. Regarding such the converged platform, S-100 standards including the international hydrographic data model are being announced and various studies have been published based on the S-100 standards. However, the existing studies have presented simple solutions for only given problems on the converged service of the maritime information. They could not propose the design concept of the converged platform which makes it possible to provide the standardized and integrated services among the geospatial data, the real-time maritime data, and the next ENC. Therefore, we propose design and implementation details of the converged service platform for the geospatial and the maritime data based on the S-100, WMS, WMTS, WPS, SOS standards. The proposed platform has advantages of supporting both the S-57 and the S-101, supporting the converged services of heterogeneous geospatial data and ENC data, supporting the real-time services of sensor data such as weather, AIS, and CCTV, and supporting the development of various kinds of maritime systems such as ECDIS, ECS, VTS based on the WebApp service. Finally, we proved the effectiveness of our proposed platform through the actual implementation of the converged service of geospatial data, S-101 data, and KWeather data.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.2
• /
• pp.15-23
• /
• 2017

The person Re-identification is the most challenging part of computer vision due to the significant changes in human pose and background clutter with occlusions. The picture from non-overlapping cameras enhance the difficulty to distinguish some person from the other. To reach a better performance match, most methods use feature selection and distance metrics separately to get discriminative representations and proper distance to describe the similarity between person and kind of ignoring some significant features. This situation has encouraged us to consider a novel method to deal with this problem. In this paper, we proposed an enhanced recurrent neural network with three-tier hierarchical network for person re-identification. Specifically, the proposed recurrent neural network (RNN) model contain an iterative expectation maximum (EM) algorithm and three-tier Hierarchical network to jointly learn both the discriminative features and metrics distance. The iterative EM algorithm can fully use of the feature extraction ability of convolutional neural network (CNN) which is in series before the RNN. By unsupervised learning, the EM framework can change the labels of the patches and train larger datasets. Through the three-tier hierarchical network, the convolutional neural network, recurrent network and pooling layer can jointly be a feature extractor to better train the network. The experimental result shows that comparing with other researchers' approaches in this field, this method also can get a competitive accuracy. The influence of different component of this method will be analyzed and evaluated in the future research.