• 한국멀티미디어학회논문지
• /
• 제9권1호
• /
• pp.11-22
• /
• 2006

얼굴 검출은 개인 인증이나 보안 등 그 응용분야가 다양하여 활발히 연구가 진행되고 있다. 본 논문에서는 분포맵에 기반한 얼굴 검출의 새로운 방법을 제안한다. 제안한 방법은 먼저, 빈도수를 고려한 피부색 분포맵을 입력 영상에 적용하여 일차적으로 얼굴 후보영역을 구하고, 이 영역에서 눈동자색 분포맵을 이용하여 눈후보영역을 결정함으로써 얼굴 구성 요소를 탐색하는 탐색공간을 축소한다. 결정된 눈 후보영역에서 가중치가 있는 윈도우를 이용하여 휘도성분과 색상성분의 상관값을 특징벡터로 한 템플릿 정합 방법으로 눈 후보점을 검출한다. 최종적으로 각 눈 후보점 쌍에 대하여 눈과 입의 위치관계 정보와 입을 인식하는 평가함수를 이용하여 얼굴을 검출하였다. 실험 결과, 제안된 방법은 좋은 성능을 보였다.

• 한국대기환경학회지
• /
• 제21권1호
• /
• pp.63-72
• /
• 2005

In order to understand regional wind characteristics and to estimate offshore wind resources, a wind map of the Korean Peninsula was established using remote-sensing data from the satellite, U.S. NASA Quik SCAT which has been deployed for the Sea Winds Project since 1999. According to the linear regression result between the wind map data and in-situ marine-buoy data, the correlation factor was greatly improved up to 0.87 by blending the remote-sensing data of Quik SCAT with U.S. NCEP/NCAR CDAS reanalysis data to eliminate precipitation interference and to increase temporal resolution. It is found from the established wind map that the wind speed in winter is prominent temporally and the South Sea shows spatially high energy density over the wind class 6. The reason is deduced that the north-west winds through the Yellow Sea and the north-east winds through the East Sea derived by the low pressure developed in Japan are accelerated passing through the Korea Channel and formed high wind energy region in the South Sea; the same trends are confirmed from the statistical analysis of the meteorological observation data of KMA.

• Membrane and Water Treatment
• /
• 제10권4호
• /
• pp.313-320
• /
• 2019

Increasing the concentration of nitrate ions in the soil solution and then leaching it to underground aquifers increases the concentration of nitrate in the water, and can cause many health and ecological problems. This study was conducted to evaluate the vulnerability of Meymeh aquifer to nitrate pollution. In this research, sampling of 10 wells was performed according to standard sampling principles and analyzed in the laboratory by spectrophotometric method, then; the nitrate concentration zonation map was drawn by using intermediate models. In the drastic model, the effective parameters for assessing the vulnerability of groundwater aquifers, including the depth of ground water, pure feeding, aquifer environment, soil type, topography slope, non-saturated area and hydraulic conductivity. Which were prepared in the form of seven layers in the ARC GIS software, and by weighting and ranking and integrating these seven layers, the final map of groundwater vulnerability to contamination was prepared. Drastic index estimated for the region between 75-128. For verification of the model, nitrate concentration data in groundwater of the region were used, which showed a relative correlation between the concentration of nitrate and the prepared version of the model. A combination of two vulnerability map and nitrate concentration zonation was provided a qualitative aquifer classification map. According to this map, most of the study areas are within safe and low risk, and only a small portion of the Meymeh Aquifer, which has a nitrate concentration of more than 50 mg / L in groundwater, is classified in a hazardous area.

• 대한수의학회지
• /
• 제60권3호
• /
• pp.117-122
• /
• 2020

Johne's disease (JD) caused by Mycobacterium avium subspecies paratuberculosis (MAP) is a chronic, wasting infectious disease in ruminants that causes enormous economic losses to the dairy and beef cattle industries. The most effective way to eradicate JD is to detect infected individuals as early as possible and remove them from the herd. However, it is difficult to detect infected individuals early with the currently using diagnostic methods. Two serological diagnostic kits commercially used worldwide and a fecal detection test were compared using 298 serum samples and feces of cattle in this study to present an efficient diagnostic method. Although there was a high correlation between the 2 serological diagnostic kits (R² = 0.7473), kit A showed a higher serological positive rate. However, the correlation between fecal tests and serological diagnosis was very low. MAP was also detected in fecal tests in many serologically negative individuals. In the periodical diagnosis of JD, MAP was detected in the feces of only cows with the higher antibody titer to MAP. These results suggest that for effective eradication of JD, early detection of infected individuals by fecal tests together with the serological tests currently in use and by removal of infected individuals are needed.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제31권7호
• /
• pp.939-948
• /
• 2004

본 논문에서는 동일 환자에 대해 시간차를 두고 촬영한 뇌 CT-CT 혈관조영영상간 움직임을 보정하기 위한 강인하고 고속의 정합방법을 제안한다. 먼저, 두 영상에서 3차원 경계검출 기법을 이용하여 특징점을 추출하고, 기준영상에서는 이를 지역적 가중치 3차원 거리맵으로 변환한다. 부유영상을 기준영상으로 강체변환하면서 두 경계간의 상관관계가 최대인 위치를 탐색한다. 이 때, 최대위치가 더 이상 변화하지 않고 일정 이상 반복되면 해당위치를 최적위치로 하여 부유영상을 최적위치로 변환시켜 두 영상을 정합한다. 실험을 위하여 인공영상을 사용하여 정화성과 강인성을 평가하였고, 육안평가를 위하여 뇌 CT-CT 혈관조영영상을 사용하였다. 본 제안방법은 지역적 가중치 3차원 거리맵을 이용함으로써 적은 샘플링 개수에도 국부최대인 위치에 수렴하지 않고 최적위치로 강인하면서 고속으로 영상이 정합되었다

• 대한원격탐사학회지
• /
• 제25권4호
• /
• pp.375-381
• /
• 2009

본 논문에서는 다양한 종류의 지표면에 대하여 분석하여 산란 특성을 연구하고 SAR 클러터 영상을 제작하고 실제 SAR 클러터 영상과 비교한다. 먼저 지표면의 특성을 분석하기 위해 각각의 지표면에 대해서 입력변수를 측정한다. 측정한 데이터를 이용하여 Oh 모델, PO 모델, radiative transfer model(RTM)을 이용하여 각도 별 산란계수를 구하였다. SAR 영상 생성을 위해 먼저 측정 지역의 DEM (digital elevation map)과 LCM (land cover map)데이터를 제작한다. DEM 데이터의 단일 픽셀(pixel)의 높이 정보를 이용하여 픽셀의 입사각을 계산하고 입사각에 따른 해당 지표면의 산란 계수를 대입한다. LCM 데이터는 해당 지역의 답사를 통해 논, 밭, 산, 길, 인공물 등을 1:5000 지도에 기입하여 SAR 영상 생성에 사용한다. DEM 데이터와 LCM 데이터를 사용하여 입사각과 지표면 종류에 따른 계수를 계산하고 영상잡음(speckle)과 영상질감(texture)을 이용하여 SAR 클러터 영상을 생성하고 실제 영상과 비교한다.

• 환경영향평가
• /
• 제22권5호
• /
• pp.455-466
• /
• 2013

The aim of this study is to seek application plans of National Ecosystem Survey database based on comparison and examination of feasible analysis methods for distribution characteristics of alien plants. In order to set up a correlation analysis method between alien plants and environmental factors, we had reviewed the 3rd National Ecosystem Survey guide book and consequently, two kinds of analysis method were tested. One was 1/25,000 scale map boundary based analysis and the other was representative mountain area based analysis. In this study we restricted the research area to select reliable surveyed database from whole "2011 National Ecosystem Survey flora database" according to two major criteria. First, an area defined by 1/25,000 scale map boundary and representative mountain area where the number of surveyed flora records should be within top 20%. Second, land cover map should also be built up inside that area. As a result, 25 map boundaries and 25 representative mountain areas were extracted to be analyzed. To limit a boundary for every representative mountain area we had analyzed distribution of environmental factors around that area by manual inspection with SPOT-5 remote sensed satellite image then designated 3km buffer zone from each alien plant location in that area. After then, naturalized index (NI) and urbanized index (UI) was calculated and correlations analysis was carried out. With the result of correlation analysis by map boundary only agricultural land area showed significant value of r (0.4~0.6, correlated) and the rest of factors did not. In the case of representative mountain area, the result showed that agricultural land, road and forest area showed significant value of r (0.6~0.8, highly correlated) which was corresponding to existing researches. Therefore, representative mountain area based method is preferable when using the alien plants database of National Ecosystem Survey for species distribution analysis. And also, considering the way of database utilization is strongly suggested at the first stage of survey planning for promoting active use of national ecosystem survey database.

• 한국지능시스템학회논문지
• /
• 제25권5호
• /
• pp.502-508
• /
• 2015

스테레오 영상은 2-D 영상으로 분석할 수 없는 깊이(거리) 정보를 포함하고 있다. 하지만 연산을 통해서 거리정보를 얻을 수 있기 때문에 계산 값의 신뢰도가 낮고, 폐색된 공간 등의 영향으로 오차가 발생한다. 또한 Stereo Matching 시 Global Method를 사용할 경우, 많은 연산량에 따라 계산 시간이 오래 걸린다. 따라서 본 논문에서는 연산 시간이 짧고 더 높은 정확도를 갖는 Disparity Map을 구하는 방법을 제안한다. 특징 기반 영상분할 기법인 윤곽선 추출을 통해 정확도는 높이고 연산 시간은 줄였다. 컬러 기반 영상 분할 기법인 Color K-Means를 통해 관심 영역을 추출하고, 이를 기반으로 Loopy Belief Propagation(LBP)을 접목하였다. 제안하는 방법을 적용함으로 영상 내 물체들의 연관성을 고려한 보정이 가능하였고, 관심 영역 추출에 따라 연산 시간을 줄일 수 있었다. 실험 결과, 기존의 방법들보다 연산 시간이 짧고 정확도가 높은 Disparity Map을 얻을 수 있었다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 하계종합학술대회 논문집(4)
• /
• pp.118-122
• /
• 2000

Nowadays, it is popular to use the spread spectrum watermarking algorithm for still image. But there is high error probability of the retrieved watermark in the spread spectrum owing to the correlation between image and spreaded watermark sequence. In this paper, two methods are proposed. One is Ordering Map Method and the other is Alteration of Image. Based on pixel value, the order by which the spreaded watermark bits is embedded is created in Ordering Map Method. By the covariance function between image and the spreaded sequence, image is altered in Alteration of Image. Hence, bit error of retrieved watermark is clearly reduced to zero by this two method.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2005년도 Proceedings of ISRS 2005
• /
• pp.687-690
• /
• 2005

Recently there have been a lot of kinds of damages in forest area such as forest fires, forest pest, landslide so that the efficient methods to mange those information and the way to face them are deadly needed. In this study, there were preparing the various vegetation index map and comparing them with the field surveying the tried to figure out which vegetation index algorism is the best proper to present forest fire damaged area. These all were based on Landsat ETM+ satellite image (2000.10.16). The result of this study is to select the high correlation algorism among the various vegetation indexes and then construct the forest fire disaster map, the case of forest fires damaged area.