• Title/Summary/Keyword: EAST 모델

Search Result 568, Processing Time 0.029 seconds

Real-time Character Detection System Using EAST Model and OCR (EAST 모델과 OCR을 이용한 실시간 문자 탐지 시스템)

  • Ye-Jun Choi;Mikyeong Moon
    • Proceedings of the Korean Society of Computer Information Conference
    • /
    • 2023.07a
    • /
    • pp.683-684
    • /
    • 2023
  • 웹페이지나 디지털 문서에는 특정 단어나 특정 문구를 검색하는 기능이 있다. 인쇄된 도서나 참고서 등과 같은 인쇄물에는 실시간으로 특정 단어나 특정 문구를 찾는 기능이 없어 어려움을 겪는 경우가 많다. 본 논문에서는 텍스트를 감지(Detection)하는 EAST 모델과 텍스트를 인식(Recognition)하는 EasyOCR을 활용한 실시간 문자 탐지 시스템의 개발내용에 대해 기술한다. 이 시스템을 통해 사용자는 인쇄물에서 실시간으로 원하는 단어나 문구를 찾아 필요한 정보를 빠르게 읽는 것에 효과적일 것을 기대한다.

  • PDF

Geoacoustic Model at the SSDP-105 Long-core Site of the Ulsan Coastal Area, the East Sea (동해 울산 연안해역 SSDP-105 심부코어 지점의 지음향 모델)

  • Ryang, Woo-Hun;Lee, Gwang-Soo;Hahn, Jooyoung
    • Journal of the Korean earth science society
    • /
    • v.39 no.2
    • /
    • pp.154-163
    • /
    • 2018
  • Geoacoustic model comprises physical and acoustic properties of submarine bottom layers influencing sound transmission through sea water and underwater. This study suggested for the first time that we made a geoacoustic model of long-coring bottom layers at the SSDP-105 drilling site of the Ulsan coastal area, which is located in the southwestern inner shelf of the East Sea. The geoacoustic model of 52 m depth below seafloor with three-layer geoacoustic units was reconstructed in the coastal sedimentary strata at 79 m in water depth. The geoacoustic model was based on the data of a deep-drilled sediment core of SSDP-105 and sparker seismic profiles in the study area. For actual modeling, the geoacoustic property values of the models were compensated to in situ depth values below the sea floor using the Hamilton modeling method. We suggest that the geoacoustic model be used for geoacoustic and underwater acoustic experiments of mid- and low-frequency reflecting on the deep bottom layers in the Ulsan coastal area of the East Sea.

A Study on the Applicability of East Asian Welfare State Discourse to the Development of Korean Welfare State(Regime) (서구적 동아시아 복지국가 담론의 한국에 대한 적용 가능성 연구)

  • Na, Byong Kyun
    • Korean Journal of Social Welfare Studies
    • /
    • v.41 no.3
    • /
    • pp.5-27
    • /
    • 2010
  • The objective of this study is to examine the dicourse of East Asian Welfare Models and its applicability to explain the development of the Korean Welfare State(regime). The Confucianist Welfare State, representing East Asian Welfare Models has limitations in explaining the current features and dynamics of east asian welfare states. Compared to the western and northern european welfare states, the Welfare State(Regime) of Korea, one of the east asian countries show the clear differences in terms of its origin, the background of formulation, the orientation of the policy and the actors. Thus, a new approach to examine the East Asian Welfare Model is needed. As a theoretical framework, the origin and the growth of the Korean Welfare State(Regime) can be understood and explained in the boundary of concept and the theoretical framework of the Authoritative Developmental State; Historically, the democratization of the 1980s and the financial crisis of the end of 1990s in Korea contributed to the growth of the welfare policies and institutions in Korea. Japan, Taiwan and other east asian counties, such as Hong Kong and Singapore have many similarities with Korea in terms of welfare policies and institutions. Comparative studies of these countries can create significant and useful results to develop a new concept and model of East Asian Welfare States. Case studies and active academic exchanges among welfare state researchers in these east asian countries are very important to develop a new concept of East Asian Welfare State Models.

A Numerical Modeling of the East sea circulation (동해 순환의 수치모델)

  • Seung, Young-Ho;Kim, Kyun
    • 한국해양학회지
    • /
    • v.28 no.4
    • /
    • pp.292-304
    • /
    • 1993
  • The east Sea circulation is numerically modeled with refined grid resolution elaborated open boundary condition, and by directly imposing the measured surface temperature and salinity typical the east Korean Warm current are clearer than those in previous works. among others, The Ulleung warm Water and the Intermediate Water of minimum salinity are nicely reproduced. The latter is formed in the northern/northwestern coastal region in winter and is advocated southward by strong under-current. the former is associated with a locally generated anti-cyclonic gyres. The model indicates strong seasonal variation of Nearshore Current along the Japanese coast from wintertime barotropic to summertime baroclinic structures. the associated strong reversed under-cur-rent in summer is not well understood. Global circulation pattern is characterized by two regions of cyclonic and anti-cyclonic gyres in the north and south, respectively. The presence of these gyres indicates importance of local dynamics in East Sea circulation. This model, however, does not completely resolve the problem of overshooting of the East Korean Warm current.

  • PDF

A Study of Machine Learning Model for Prediction of Swelling Waves Occurrence on East Sea (동해안 너울성 파도 예측을 위한 머신러닝 모델 연구)

  • Kang, Donghoon;Oh, Sejong
    • The Journal of Korean Institute of Information Technology
    • /
    • v.17 no.9
    • /
    • pp.11-17
    • /
    • 2019
  • In recent years, damage and loss of life and property have been occurred frequently due to swelling waves in the East Sea. Swelling waves are not easy to predict because they are caused by various factors. In this research, we build a model for predicting the swelling waves occurrence in the East Coast of Korea using machine learning technique. We collect historical data of unloading interruption in the Pohang Port, and collect air pressure, wind speed, direction, water temperature data of the offshore Pohang Port. We select important variables for prediction, and test various machine learning prediction algorithms. As a result, tide level, water temperature, and air pressure were selected, and Random Forest model produced best performance. We confirm that Random Forest model shows best performance and it produces 88.86% of accuracy

Geoacoustic Model of Coastal Bottom Strata at Jeongdongjin in the Korean Continental Margin of the East Sea (동해 한국대륙주변부 정동진 연안 지층의 지음향 모델)

  • Ryang, Woo-Hun;Kim, Seong-Pil;Kim, Dae-Choul;Hahn, Jooyoung
    • Journal of the Korean earth science society
    • /
    • v.37 no.4
    • /
    • pp.200-210
    • /
    • 2016
  • Geoacoustic modeling is used to predict sound transmission through submarine bottom layers of sedimentary strata and acoustic basement. This study reconstructed four geoacoustic models for sediments of 50 m thick at the Jeongdongjin area in the western continental margin of the East Sea. Bottom models were based on the data of the highresolution air-gun seismic and subbottom profiles (SBP) with sediment cores. P-wave speed was measured by the pulse transmission technique, and the resonance frequency of piezoelectric transducers was maintained at 1MHz. Measurements of 42 P-wave speeds and 41 attenuations were fulfilled in three core sediments. For actual modeling, the P-wave speeds of the models were compensated to in situ depth below the sea floor using the Hamilton method. These geoacoustic models of coastal bottom strata will be used for geoacoustic and underwater acoustic experiments reflecting vertical and lateral variability of geoacoustic properties in the Jeongdongjin area of the East Sea.

Performance Improvement Analysis of Building Extraction Deep Learning Model Based on UNet Using Transfer Learning at Different Learning Rates (전이학습을 이용한 UNet 기반 건물 추출 딥러닝 모델의 학습률에 따른 성능 향상 분석)

  • Chul-Soo Ye;Young-Man Ahn;Tae-Woong Baek;Kyung-Tae Kim
    • Korean Journal of Remote Sensing
    • /
    • v.39 no.5_4
    • /
    • pp.1111-1123
    • /
    • 2023
  • In recent times, semantic image segmentation methods using deep learning models have been widely used for monitoring changes in surface attributes using remote sensing imagery. To enhance the performance of various UNet-based deep learning models, including the prominent UNet model, it is imperative to have a sufficiently large training dataset. However, enlarging the training dataset not only escalates the hardware requirements for processing but also significantly increases the time required for training. To address these issues, transfer learning is used as an effective approach, enabling performance improvement of models even in the absence of massive training datasets. In this paper we present three transfer learning models, UNet-ResNet50, UNet-VGG19, and CBAM-DRUNet-VGG19, which are combined with the representative pretrained models of VGG19 model and ResNet50 model. We applied these models to building extraction tasks and analyzed the accuracy improvements resulting from the application of transfer learning. Considering the substantial impact of learning rate on the performance of deep learning models, we also analyzed performance variations of each model based on different learning rate settings. We employed three datasets, namely Kompsat-3A dataset, WHU dataset, and INRIA dataset for evaluating the performance of building extraction results. The average accuracy improvements for the three dataset types, in comparison to the UNet model, were 5.1% for the UNet-ResNet50 model, while both UNet-VGG19 and CBAM-DRUNet-VGG19 models achieved a 7.2% improvement.

Mathematical Modelling of Tides and Surges in the East China Sea (동지나해의 조석 및 해일 수치모델)

  • 최병호
    • Water for future
    • /
    • v.16 no.4
    • /
    • pp.221-236
    • /
    • 1983
  • In semi-enclosed shallow sea areas typified by the Yellow sea and the East China Sea, currents and sea surface variations are predominantly tidal. During the recent years two-dimensional numerical hydrodynamic model of the Yellow Sea and the East China Sea has been developed, based on the vertically-integrated equations of motion and continuity, capable of reproducing amplitudes and phases of the principal components of tides to satisfiable accuracy. As a subsequent development a three-dimensional hydrodynamical nymerical model covering the Yellow Sea and the East China Sea has been formulated to investigate the vertical distribution of horizontal tidal current and the response of the continented to investigate the vertical distribution of horizontal tidal current and the response of the continental shelf sea to steady uniform wind stress field imposed over the surface. Features of the M2 tidal current and the wind-induced three-dimensional current structure determined from the computation have been examined and discussed.

  • PDF

Atmospheric and Oceanic Factors Affecting the Air-Sea Thermal Interactions in the East Sea (Japan Sea) (東海海面 熱交換에 影響을 미치는 大氣 및 海洋的 要因)

  • Kang, Yong Q
    • 한국해양학회지
    • /
    • v.19 no.2
    • /
    • pp.163-171
    • /
    • 1984
  • The atmospheric and oceanic influences on the air-sea thermal interaction in the East Sea (Japan Sea) are studied by means of an analytic model which is based on the heat budget of the ocean. By means of the model, the model, the annual variations of heat fluxes and air temperatures in the East Sea are analytically simulated. The model shows that the back radiation, the latent heat and the sensible heat increase with the warn water advection. The latent heat increases with the sea surface temperature (SST) but the back radiation and the sensible heat dcrease as the SST increases. In the East Sea, an increase of mean SST by 1.0$^{\circ}C$ yields an increase of mean air temperature by 1.2$^{\circ}C$. The heat storage in the ocean plays an important role in the annual variations of heat flux across the sea surface.

  • PDF

Neuroprotecticve Effect of Rhei Rhizoma on Transient Global Ischemia in Gerbil (Gerbil의 전뇌허혈에 대한 대황의 신경보호효과)

  • Bum-Hoi, Kim;Hyuk-Sang, Jung;Ran, Won;Ji-Ho, Park;Chul-Hun, Kang;Nak-Won, Sohn
    • The Journal of Korean Medicine
    • /
    • v.23 no.3
    • /
    • pp.74-84
    • /
    • 2002
  • 목적 : 본 실험에서는 gerbil을 이용한 전뇌허혈 동물모델에서 뇌허혈손상 직후 지연성 뇌손상에 대한 대황의 방어효과와 Apoptosis 과정중의 Bax와 Bcl-2 단백질에 대한 조절작용을 관찰하고, TUNEL 염색법을 통하여 대황이 gerbil hippocampus CAl영역의 pyramidal neuron의 세포사에 미치는 영향과 PCl2세포를 이용한 세포배양 모델에서의 대황의 신경방어 효과를 관찰하였다. 방법 : Mongolian gerbil의 총경동맥을 5분간 폐색하여 가역성 전뇌허혈을 유발시킨 후 대황의 전탕액을 하루에 한번 경구 투여하였다. 대황의 신경 보호 효과는 수술 7일 후에 cresyl violet으로 염색하여, 살아있는 신경 세포의 수를 세어 측정하였다. 또, 수술 3일 후에는 면역조직화학적 방범을 통하여 Bax. Bcl-2단백질의 발현과 대황의 신경보호 효과와의 관련성을 알아보았다. 결과: 가역적 전뇌허혈이 일어난 동물군의 경우 hippocampus의 CAl 영역에서 살아있는 신경세포의 수는 $51.0{\pm}2.5개{\;}/mm$에 불과하였으나, 그에 비해 수술 후 7일간 대황을 투여한 동물군은 $106.2{\pm}2.5개{\;}/mm$로 살아 있는 신경세포수가 크게 증가하였다. Apoptosis를 촉진하는 단백질인 Bax의 발현은 3일간 대황을 투여한 동물군의 경우 hippocampus의 CAl 영역에서 현저하게 저해되었고, Apoptosis를 억제하는 Bcl-2 단백질의 발현은 변화가 없었다. TUNEL assay를 통하여 살펴본 결과 대황 투여군의 apoptotic 신경세포사가 감소하였으며 이는 Bax protein의 발현과 유사한 양상을 나타내었다. 결론:대황이 Bax 단백질의 발현을 억제하여 상대적으로 Bax/Bcl-2 자연적 세포사를 억제하여 Mogolian gerbil의 가역성 전뇌허혈 모델에서 신경보호효과를 나타내는 것으로 사료된다.

  • PDF