• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.6
• /
• pp.227-233
• /
• 2019

Recently, along with the recent development of deep learning technique, neural networks are achieving success in computer vision filed. Convolutional neural network have shown outstanding performance in not only for a simple image classification task, but also for tasks with high difficulty such as object segmentation and detection. However many such deep learning models are based on supervised-learning, which requires more annotation labels than image-level label. Especially image semantic segmentation model requires pixel-level annotations for training, which is very. To solve these problems, this paper proposes a weakly-supervised semantic segmentation method which requires only image level label to train network. Existing weakly-supervised learning methods have limitations in detecting only specific area of object. In this paper, on the other hand, we use multi-classifier deep learning architecture so that our model recognizes more different parts of objects. The proposed method is evaluated using VOC 2012 validation dataset.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.9 s.351
• /
• pp.128-134
• /
• 2006

Traffic information protocol in DMB is very different from existing analog broadcasting and wireless communication network. In this paper, we examined whether traffic information protocol of Europe Broadcasting Union(TPEG) is applicable to domestic DMB. Also, we proposed a division of classification on kinds of franc information, and related data that it is required to transmit traffic information of TPEG form. We composed of experiment equipment and studied whether is expressed traffic informations as like accident, event, traffic condition and CCTV image on car navigation system. The results obtained it can be given expression to phrases from TPEG streaming data and to link with electronics map by decoding TPEG straming data. Also it can be expressed CCTV and graphic image which is composed of TPEG form.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2003.11a
• /
• pp.410-415
• /
• 2003

Ground subsidence has occurred in the downtown of Muan-eup in Korea. Integrated geophysical survey, including two-dimensional resistivity, CSMT(Controlled source magnetotelluric), magnetic, borehole logging, GPR and resistivity tomography, has been conducted to investigate the cause of subsidence and ground conditions. Since the target area is in the city downtown, there were no spaces for surface geophysical methods. To get regional geology and to facilitate the detailed geophysical interpretation in the survey area, two-dimensional resistivity, CSMT and magnetic surveys have been applied in the outer region of the downtown. From these results, we could accurately define the Gwangju fault system and estimate the geologic conditions in the downtown. For the detailed survey of the downtown area, resistivity tomography and borehole logging data have been acquired using a few tens of densely located boreholes. Among these survey results, borehole logging data provided the guide to classification of the rock type and we could define the geologic boundary of granite and limestone formations. From the resistivity tomograms of 42 sections, which are densely located enough to be interpreted in a three-dimensional manner, we could delineate the possible weak zones or cavities in the limestone formations. In particular, resistivity tomograms in the subsided area showed the real image of ground subsidence. The map of hazardous zone has been derived from the joint interpretation of these survey results and we could provide the possible reinforcement strategy in this area.

• Journal of Ginseng Research
• /
• v.42 no.1
• /
• pp.57-67
• /
• 2018

Background: Ginseng contains many small metabolites such as amino acids, fatty acids, carbohydrates, and ginsenosides. However, little is known about the relationships between microorganisms and metabolites during the entire ginseng fermentation process. We investigated metabolic changes during ginseng fermentation according to the inoculation of food-compatible microorganisms. Methods: Gas chromatography mass spectrometry (GC-MS) datasets coupled with the multivariate statistical method for the purpose of latent-information extraction and sample classification were used for the evaluation of ginseng fermentation. Four different starter cultures (Saccharomyces bayanus, Bacillus subtilis, Lactobacillus plantarum, and Leuconostoc mesenteroide) were used for the ginseng extract fermentation. Results: The principal component analysis score plot and heat map showed a clear separation between ginseng extracts fermented with S. bayanus and other strains. The highest levels of fructose, maltose, and galactose in the ginseng extracts were found in ginseng extracts fermented with B. subtilis. The levels of succinic acid and malic acid in the ginseng extract fermented with S. bayanus as well as the levels of lactic acid, malonic acid, and hydroxypruvic acid in the ginseng extract fermented with lactic acid bacteria (L. plantarum and L. mesenteroide) were the highest. In the results of taste features analysis using an electronic tongue, the ginseng extracts fermented with lactic acid bacteria were significantly distinguished from other groups by a high index of sour taste probably due to high lactic acid contents. Conclusion: These results suggest that a metabolomics approach based on GC-MS can be a useful tool to understand ginseng fermentation and evaluate the fermentative characteristics of starter cultures.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.189-193
• /
• 2011

유역관리 및 수질 향상을 위해 다양한 환경정책이 시행되고 있으며, 최근 수질오염총량관리제의 시행으로 인해 보다 집중적인 유역관리와 수질 향상을 위한 노력이 배가되고 있다. 이러한 노력의 일환으로 현재 환경부 국립환경과학원에서는 수질오염총량관리를 위하여 단위유역의 말단지점에서 수질 및 유량자료에 대한 정기적인 측정을 8일 간격으로 시행하고 있으며, 데이터 베이스 및 웹시스템을 통하여 자료를 공개하고 있다(이호열, 2009). 이와 같은 자료의 측정과 축적은 그 분석을 통해 수질 개선을 위한 우선 관리 대상지점의 파악 등과 같이 수질오염총량관리제의 시행과 평가를 위해 사용될 수 있을 뿐만 아니라 새로운 환경정책의 수립에도 활용될 수 있을 것으로 기대된다. 그러나 현재 소수의 연구들에서만 상기의 자료를 단순히 활용한 결과를 찾을 수 있으며, 특히 측정된 수질 및 유량자료를 분석하여 발표한 연구결과 역시 소수에 지나지 않는다(김철겸 등, 2009). 측정 자료에 대한 분석 및 이에 따른 자료의 활용성 제고를 위해서 다양한 자료 분석 기법의 개발과 적용이 절실하다. 이러한 자료 분석 기법의 개발 및 적용에 관한 연구의 일환으로 최근 패턴분류를 위해 다양한 분야에서 활용되고 있는 자기조직화 특성 지도(Self Organizing Feature Map: SOFM)를 상기의 측정 자료에 적용한 연구 결과가 보고된 바 있다(진영훈 등, 2009; 2010). 본 연구에서는 수질오염총량관리제를 위해 측정되고 있는 수질 및 유량자료를 수집하여 자료에 내재되어 있는 시 공간적 특성을 분석하고자 하였다. 영산강 유역을 대상으로 하여, 본 유역 내의 단위유역들 중 황룡_A, 지석_A, 영본_A, 영본_B, 영본_C, 영본_D의 말단지점에서 측정되고 있는 BOD (Biochemical Oxygen Demand), TOC (Total Organic Carbon), T-N (Total Nitrogen), T-P (Total Phosphorus), SS (Suspended Solids) 수질농도 및 유량자료를 대상으로 연구를 진행하였다.

• Journal of Korea Water Resources Association
• /
• v.51 no.11
• /
• pp.1011-1020
• /
• 2018

This study aims to develop a drought monitoring scheme based on groundwater which can be exploit for water supply under drought stress. In this context, groundwater level can be used as a proxy for better understanding the temporal evolution of drought state. First, kernel density estimator is presented in the monthly groundwater level over the entire national groundwater stations. The estimated cumulative distribution function is then utilized to map the monthly groundwater level into the standardized groundwater level index (SGI). The SGI for each station was eventually converted into the index for major cities through the Thiessen polygon approach. We provide a drought classification for a given SGI to better characterize the degree of drought condition. Ultimately, we conclude that the proposed monitoring framework enables a more reliable estimation of the drought stress, especially for a limited water supply area.

• Ecology and Resilient Infrastructure
• /
• v.5 no.2
• /
• pp.59-71
• /
• 2018

Lake Enriquillo, the largest lake in the Dominican Republic, recently has undergone unusual water area changes since 2001 thus it has been affected seriously by local community's livelihood. Earthquakes and seismic activities of Hispaniola plate tectonic coupled with human activities and climate change are addressed as factors causing the increasing. Thus, a thorough study on relationship between lake area changing, and those factors is needed urgently. To do so, this study applied MESMA on MODIS data to extract water area of Lake Enriquillo during 2001 and 2012 bimonthly, with six issues 12-year. MODIS provides high temporal resolution, and its coarse spatial resolution is compensated by MESMA fraction map. The increase in water area was $142.2km^2$, and the maximum lake area was $338.0km^2$ (in 2012). Water areas extracted by two Landsat scenes at two different times with three image classification approaches (ISODATA, MNDWI, and TCW) were used to assess accuracy of MODIS and MESMA results; it indicated that MESMA water areas are same as ISODATA's, less than 0.4%, while the highest difference is between MESMA and TCW, 2.4%. A number of previously formulated hypotheses of lake area change were investigated based on the outcomes of the present study, though none of them could fully explain the changes.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.52 no.3
• /
• pp.39-45
• /
• 2010

Quantification of carbon absorption and understanding the human induced land use changes (LUC) forms one of the major study with respect to global climatic changes. An attempt study has been made to quantify the carbon absorption by LUC through remote sensing technology. The Landsat imagery four time periods was classified with the hybrid classification method in order to quantify carbon absorption by LUC. Thereafter, for estimating the amount of carbon absorption, the stand biomass of forest was estimated with the total weight, which was the sum of individual tree weight. Individual tree volumes could be estimated with the crown width extracted from digital forest cover type map. In particular, the carbon conversion index and the ratio of the $CO_2$ molecular weight to the C atomic weight, reported in the IPCC guideline, was used to convert the stand biomass into the amount of carbon absorption. Total carbon absorption has been modeled by taking areal estimates of LUC of four time periods and carbon factors for land use type and standing biomass. Results of this study, through LUC suggests that over a period of construction, 7.10 % of forest and 9.43 % of barren were converted into urban. In the conversion process, there has been a loss of 6.66 t/ha/y (7.94 %) of carbon absorption from the study area.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2007.05a
• /
• pp.1322-1326
• /
• 2007

지구 온난화에 의한 엘리뇨, 라니냐의 발생은 한반도에 커다란 기후변화를 가져왔다. 이러한 변화는 연간 총 강우량은 크게 변화하지 않았지만 강우강도가 증가하고, 태풍의 규모가 커지는 등 홍수방어의 측면에서 상당한 어려움을 야기하고 있다. 2006년 태풍 '에위니아'와 이에 뒤이은 태풍과 중국 북부지방의 기단이 장마전선을 형성하며 사상 최고의 강우강도를 기록한 집중호우는 미래의 기후조건이 극단의 사상으로 변화하고 있으며 이에 따른 구조적/비구조적 대책이 요구됨을 단적으로 드러내는 예이다. 이러한 집중호우들의 결과로 발생하는 제방붕괴 등으로 인한 홍수범람은 수많은 인명과 재산의 피해를 발생시키고 있다. 따라서, 본 연구에서는 500년 빈도 홍수유출에 의한 제방붕괴모의를 실시하고 강원도 원주시에 대한 침수심과 범람범위를 도출하고, 해당 침수지역의 홍수위험강도(침수심${\times}$유속)를 구하였고, FEMA에서 제시한 바 있는 Downstream Hazard Classification Guidelines(USBR, 1988)의 사람 및 가옥에 대한 위험 정도를 도출하였으며, 이 자료들을 근간으로 하여 침수심과 범람범위의 정보를 포함하고 있던 범람지도와 지역별 홍수의 세기를 표현하는 홍수위험 정보를 GIS Tool(ArcView, ArcGIS, CAD 등)들을 이용하여 표출함으로써 좀 더 효율적이고 체계적인 홍수지도를 작성하였다. 본 연구의 결과로 작성된 홍수위험정보 지도(침수심, 유속분포도, 홍수위험강도)는 주민의 대피 이동시나 대피지구 선정, 홍수로부터 발생하는 피해에 대한 최적화된 대응 대비, 도시개발계획 수립시 개발지역의 선정 등에 기초자료로 사용될 수 있을 것으로 판단되었다. 다양한 정보를 포함하고 있는 홍수정보 지도의 작성을 통해 기존 침수깊이와 범람범위의 1차적인 정보에서 수립할 수 있었던 계획이 가질 수 있었던 한계점을 좀 더 향상된 정보를 이용하여 극복함으로써 효율적이고 체계적인 치수 방재 계획수립이 가능할 것으로 판단된다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.3
• /
• pp.209-214
• /
• 2010

This paper proposes an algorithm of segmentation and title detection for document image. The automated title detection method that we have developed is composed of two phases, segmentation and title area detection. In the first phase, we extract and segment the document image. To perform this operation, the binary map is segmented by combination of morphological operation and CCA(connected component algorithm). The first phase provides segmented regions that would be detected as title area for the second stage. Candidate title areas are detected using geometric information, then we can extract the title region that is performed by removing non-title regions. After classification step that removes non-text regions, projection is performed to detect a title region. From the fact that usually the largest font is used for the title in the document, horizontal projection is performed within text areas. In this paper, we proposed a method of segmentation and title detection for various forms of document images using geometric features and projection profile analysis. The proposed system is expected to have various applications, such as document title recognition, multimedia data searching, real-time image processing and so on.