• Journal of Internet Computing and Services
• /
• v.7 no.4
• /
• pp.67-77
• /
• 2006

Wrappers play an important role in extracting specified information from various sources. Wrapper rules by which information is extracted are often created from the domain-specific knowledge. Domain-specific knowledge helps recognizing the meaning the text representing various entities and values and detecting their formats However, such domain knowledge becomes powerless when value-representing data are not labeled with appropriate textual descriptions or there is nothing but a hyper link when certain text labels or values are expected. In order to alleviate these problems, we propose a probabilistic method for recognizing the entity type, i.e. generating wrapper rules, when there is no label associated with value-representing text. In addition, we have devised a method for using the information reachable by following hyperlinks when textual data are not immediately available on the target web page. Our experimental work shows that the proposed methods help increasing precision of the resulting wrapper, particularly extracting the title information, the most important entity on a web page. The proposed methods can be useful in making a more efficient and correct information extraction system for various sources of information without user intervention.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.5 no.2
• /
• pp.37-46
• /
• 1987

The single photo method has a lot of applications in forestry, traffic accident managements, industry, criminal investigation, and in daily life. In this study a new single photo method was developed by classifying into the Space resection method and the 2 Dimensional Perspective Transformation method. Metric and nonmetric cameras were used to analyse the accuracy by means of single photo method, and the errors in coordinates and lengths were studied by changing the number and arrangement of control points to obtain the optimum condition for the single photo method. The influence of number and arrangement of control points on the accuracy was relatively small in case of the Metric WILD P31 and ASAHI PENTAX 6$\times$7 cameras, where as for errors it was a major factor in the Non-metric Nikon FM2. To overcome these defects, at least 6 control points should be used for the errors to be convergent and they should be distributed evenly over the surveying area. It was found that accuracy increased as the object to be photographed was placed in the perpendicular direction to the axis of camera.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.5
• /
• pp.199-207
• /
• 2013

Spin images representing efficiently surface features of 3D mesh models have been used to detect facial landmark points. However, at a certain point, different normal direction can lead to quite different spin images. Moreover, since 3D points are projected to the 2D (${\alpha}-{\beta}$) space during spin image generation, surface features cannot be described clearly. In this paper, we present a method to detect 3D facial landmark using improved spin images by partitioning the search area with respect to angle. By generating sub-spin images for angular partitioned 3D spaces, more unique features describing corresponding surfaces can be obtained, and improve the performance of landmark detection. In order to generate spin images robust to inaccurate surface normal direction, we utilize on averaging surface normal with its neighboring normal vectors. The experimental results show that the proposed method increases the accuracy in landmark detection by about 34% over a conventional method.

• Journal of Korea Water Resources Association
• /
• v.46 no.10
• /
• pp.969-976
• /
• 2013

Leakage in water distribution system causes social and economic losses by direct water loss into the ground, and additional energy demand for water supply. This research suggests a leak detection model of using adaptive Kalman filtering on real-time data of pipe flow. The proposed model takes into account hourly and daily variations of water demand. In addition, the model's prediction accuracy is improved by automatically calibrating the covariance of noise through innovation sequence. The adaptive Kalman filtering shows more accurate result than the existing Kalman method for virtual sine flow data. Then, the model is applied to data from two real district metered area in JE city. It is expected that the proposed model can be an effective tool for operating water supply system through detecting burst leakage and abnormal water usage.

• KSBB Journal
• /
• v.8 no.4
• /
• pp.328-335
• /
• 1993

A flow injection analysis(FIA) system was developed for the determination of cholesterol using immobilized cholesterol oxidase and cholesterol ester hydrolase. The enzymes were immobilized on controlled pore galas(CPG) by the glutaraldehyde method. The glass colunms packed with immobilized enzymes were found to contain 3-5 I.U. for each enzyme. A hydrogen peroxide sensitive electrode was contructed and applied to the FIA system. The operational conditions for FIA response were investigated and optimized with variation of sampling volume, flow rate and composition of carrier solution. The FIA response were linear upto 60 and 400mg/m1 for free cholesterol and cholesterol ester, respectively. All samples were analyzed with a good precision (<2.5% CV) and accuracy. 23 samples were mea sured succesively within about an hour. Intermittent assays of more than 500 times caused 50% decrease in response sensitivity.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.7 no.4
• /
• pp.139-146
• /
• 1987

The 3-dimensional deformation measurement by the terrestrial photogrammetry is consist of 3-dimensional coordinates computation, displaced point detection and deformation estimation of object targets. In this study, at the first step of deformation analysis, the variation of the variance-covariance matrix for the exterior orientation elements was analyzed by the increment of the ground control points and the photos in the Bundle adjustment. And then, to give the constraints for improving accuracy of ground control points, the concept of Free-Network adjustment was applied to Bundle adjustment. As a result, we knew that it was desired in the accuracy and the economy, the observation time when the numbers of ground control point and photo were respectively 6 points and 3 photos. In addition, in the case of applying the concept of Free Network adjustment in Bundle adjutment, it was desirable that the space distance for the constraints is distributed outside.

• Journal of Korea Water Resources Association
• /
• v.42 no.3
• /
• pp.259-270
• /
• 2009

The applicability of the developed distributed rainfall runoff model using a multi-directional flow allocation algorithm and a real-time updating algorithm was evaluated. The rainfall runoff processes were simulated for the events of the Andong dam basin and the Namgang dam basin using raingauge network data and weather radar rainfall data, respectively. Model parameters of the basins were estimated using previous storm event then those parameters were applied to a current storm event. The physical propriety of the multi-directional flow allocation algorithm for flow routing was validated by presenting the result of flow grouping for the Andong dam basin. Results demonstrated that the developed model has efficiency of simulation time with maintaining accuracy by applying the multi-directional flow allocation algorithm and it can obtain more accurate results by applying the real-time updating algorithm. In this study, we demonstrated the applicability of a distributed rainfall runoff model for the advanced basin-wide flood management.

• Journal of Broadcast Engineering
• /
• v.8 no.2
• /
• pp.198-203
• /
• 2003

We develop the method of estimating the endpoints of speech by jointly using the lip motion (visual speech) and speech being included in multimedia data and then propose a new speech recognition system (SRS) based on that method. The endpoints of noisy speech are estimated as follows : For each test word, two kinds of endpoints are detected from visual speech and clean speech, respectively Their difference is made and then added to the endpoints of visual speech to estimate those for noisy speech. This estimation method for endpoints (i.e. speech interval) is applied to form a new SRS. The SRS differs from the convention alone in that each word model in the recognizer is provided an interval of speech not Identical but estimated respectively for the corresponding word. Simulation results show that the proposed method enables the endpoints to be accurately estimated regardless of the amount of noise and consequently achieves 8 o/o improvement in recognition rate.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.377-377
• /
• 2019

본 연구에서는 LSTM 모형을 이용하여 갈수예보를 위한 월 단위 전망모형개발의 대상지점으로 이수 및 치수의 측면에서 아주 중요한 한강대교 지점을 선정하였으며 유량예보를 위하여 한강수계 19개 기상관측소의 월평균강수량, 월평균기온 및 3개 댐(소양,횡성,충주)의 월방류량을 사용하여 한강대교의 월 유량을 예측하였다. 1996년부터 2016년까지의 자료는 모형의 학습, 2017년 자료는 모형의 검증에 활용하였으며 가장 최근 건설된 횡성댐 방류량의 경우 1996년~2000년의 자료가 없으므로 2001년~2005년의 자료를 반복하여 학습에 활용하였다. 모형의 예측결과는 신경망 학습 시 한강대교 월유량자료를 포함한 결과와 미포함 결과를 도출하였으며, 모의결과의 재현성 분석을 위하여 월별 예측값과 실측값의 비율을 산정하였으며 1월부터 12월까지 12개 값을 평균하여 평균예측률을 산정하고 이를 홍수기(6월~10월) 및 비홍수기(1월~5월, 11월~12월)를 구분하였다. 딥러닝 학습 시 월유량을 포함한 경우의 예측결과가 학습 시 월유량을 포함하지 않았을 경우보다 상대적으로 좋은 정확도를 보이는 것으로 분석되었다. 다만, 신경망을 실제 갈수예보에 활용하기 위해서는 예측 기상정보인 월강우량, 월평균기온, 댐방류량만을 활용하여야 하는데 학습 시월유량 미포함 결과는 예측률이 매우 낮았으며, 신경망의 학습횟수가 늘어날 경우 학습자료 과적합(over-fitting)되어 정확도가 보다 저하되는 것으로 나타났다. 그래서 기존의 현재시간 t까지의 입력자료로 학습 후 익월(t+1)의 월유량을 예측하는 (t $\rightarrow$ t+1) 방법에서 현재시점 (t-n ~ t)까지의 입력자료를 이용하여 당월(t)의 월유량을 산정하는 (t$\rightarrow$t) 방법으로 재학습 후 모형검증을 수행한 결과 전술한 익월(t+1) 유량을 예측한 결과보다 재현성이 훨씬 향상된 것으로 분석되며평균예측률이 0.99로 홍수기 및 비홍수기에서도 뛰어난 정확성을 보이고 있다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.89-89
• /
• 2019

수환경으로 유출되는 유해화학물질은 독성을 가지고 직접 유출되거나 다양한 매체와 반응하여 화재 및 폭발 등의 사고가 발생한다. 실제로 낙동강 유역에서는 1991년 페놀 유출사고를 시작으로 2009년 구미공단 '1,4-다이옥산' 유출사고, 2014년 11월 경북 봉화군의 황산유출사고 등 크고 작은 사고가 빈번히 발생하고 있으며 작년 6월에는 대구와 부산의 수돗물에서 과불화화합물이 검출되기도 하였다. 이러한 대규모 사고를 방지하기 위해 신속한 오염물의 거동 예측이 가능한 추적모델이 필요하며, 본 연구에서는 수환경으로 유출된 유해화학물질의 추적을 위한 1차원 저장대 모형을 개발하였다. 일반적으로 저장대 모형은 복잡한 하천 구조를 하천의 주 흐름이 존재하는 본류대와 하천 흐름이 정체되는 저장대, 그리고 하상구조로 단순화 하여 나타낸다. 본류대에서는 하천흐름에 의한 이송 및 횡방향 유속차로 발생하는 전단류에 의한 확산이 일어나며, 저장대와의 물질교환으로 발생하는 저장효과와, 하상구조와의 흡착 및 탈착, 그리고 생물화학적 반응 및 휘발이 발생한다고 가정한다. 본류대와 저장대간의 질량교환은 난류유속변동과 농도차에 의해서만 발생한다고 가정하고 오염물질의 이송과 분산과정을 해석한다. 저장대에서는 이송 및 전단류에 의한 확산은 일어나지 않으며, 본류대와의 물질교환으로 발생하는 저장효과와 하상구조로의 흡착, 그리고 생물화학적 반응 및 휘발이 발생한다고 가정하며, 하상구조에서는 본류대 및 저장대와의 흡착 및 탈착만 발생한다고 가정한다. 저장대 모형의 해석을 위해서는 리치(Reach) 별로 본류대 분산계수($K_F$), 본류대 면적($A_F$), 저장대 면적($A_S$), 그리고 저장대 교환계수(${\alpha}$)의 네 가지 저장대 매개변수가 필요하며 본 연구에서 개발된 저장대 모형은 흡탈착, 생물화학적 반응 및 휘발 과정을 모두 고려하여 유해화학물질의 확산 거동을 모의한다. 최적의 리치길이, 흡탈착, 반응 및 휘발 계수를 산정하여 모형의 정확도를 향상시켰으며, 신속하고 정확하게 오염물의 거동을 예측할 수 있었다.