• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.320-321
• /
• 2022

부두 내 컨테이너를 적재하는 과정에서 정렬 상태가 부정확한 경우 강풍으로 인한 안전사고가 발생할 가능성이 있다. 본 논문에서는 컨테이너 안전사고를 예방하기 위한 딥러닝 기반의 컨테이너 정렬 상태 분류 알고리즘을 제안한다. 제안하는 알고리즘은 정렬을 분류하는 기준을 제시하고 YOLO 기반의 모델을 구현했다. 추론 속도, 검출 정확도, 분류 정확도를 기준으로 각 모델의 성능을 평가했으며 성능 결과는 YOLOv4모델이 YOLOv3모델에 비해서 추론 속도는 느리지만, 검출 정확도와 분류 정확도는 높음을 보인다.

• Journal of Energy Engineering
• /
• v.4 no.1
• /
• pp.23-30
• /
• 1995

분류층 가스화기 설계를 위한 일차연구로서 가스화기 종횡비, 주입方法, 선회강도 및 주입속도 등에 따른 비반응 난류장 특성을 수치해석적 방법에 의해 파악하였다. 수치해석은 검사체적에 기초한 Patankar의 유한차분방법을 이용하였으며 압력과 속도의 연계문제는 SIMPLEC 알고리즘을, 레이놀즈 전단력은 K- 난류모델을 사용하였다. 입자궤적 계산은 공기역학적 향력만을 고려하였으며 비선형적인 공기저항력에 의한 난류변동상관모델은 고려치 않았다. 이차공기 주입방법(parallell injection과 nonparallel 3$0^{\circ}C$ imjection)에 따른 수치해석을 수행하여 Ar tracer의 질량분율 및 기타 속도에 대한 實驗資料와 비교하여 만족할 만한 結果를 얻었다. 나아가서 假想的인 가스화기 모델을 대상으로 가스화기의 종횡비, 선회강도, 주입속도 및 주입각 등에 따른 와류 形成 위치 등을 포함한 유동장 특성 및 입도에 따른 궤적분석을 시도하였다.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.4
• /
• pp.912-919
• /
• 1990

본 연구에서는 다중선회연소기의 원리에 기초하여 속도차가 있는 두 공기류의 전단층에 기체연료를 분출하여 연소시키면 연료가 두 공기류 사이에 유입되므로 연료 가 산화제의 접촉면적이 증대되고 또한 난류혼합속도가 큰 영역으로 연료가 유입되므 로 혼합효과가 증대되어 고부하연소에 적절한 방식이 될 것으로 생각하여 동축이중공 기분류중의 난류 확산화염에 대해 그 화염구조를 밝히고 이 화염을 실용연소기에 응용 하기 위한 기초자료를 얻는데 목적이 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10a
• /
• pp.207-209
• /
• 1999

클러스터 분석은 관측의 대상이 되는 집합에 맞는 분류 구조를 생성하는데 이용되는 통계학적인 기술이다. 정보검색 응용에서 전형적으로 발견되는 높은 차원을 가진 많은 데이터 집합을 클러스터하기 위하여, 많은 공간과 시간이 필요하다. SLINK 알고리즘은 O(n2)의 시간과 O(n)의 공간의 성능을 갖으며 점진성을 반영할 수 있는 알고리즘이다. SLINK알고리즘을 이용하여 검색 엔진의 검색결과에 온라인으로 클러스터 분류를 수행하는 시스템을 구현하였다. 구현된 시스템은 상대적으로 높은 정확도와 각 클러스터를 저장하고 표현하는데 있어서의 장점을 제공하며, 상대적으로 느린 수행 속도는 온라인으로 문서들이 다운로드 되는 속도가 느리므로 문제가 되지 않음을 알 수 있었다.

• Tunnel and Underground Space
• /
• v.27 no.4
• /
• pp.230-242
• /
• 2017

Seismic velocities measured from in-situ tests (n=177) and through rock core samples (n=1,035) are reviewed in light of construction standards, widely used standards as a first-hand approximation of rock classification solely based on seismic velocities. In-situ down hole tests and refraction survey for soft rocks showed seismic velocities of 1,400~2,900 m/s which is faster than those specified in construction standards. For moderate~ hard rocks, in-situ down hole tests and refraction survey showed 2,300~3,800 m/s which roughly corresponds with the range specified in the construction standards. A similar trend is also observed for seismic velocities measured from rock core samples. The observed differences between construction standards and seismic velocities can be explained in two ways. If construction standards are correct the observed differences may be explained with seismic velocities affected by underlying fast velocities and also possibly with selection of intact cores for velocity measurement. Alternatively, construction standards may have intrinsic problems, namely artificial discrete boundaries between soft rocks and moderate rocks, application of foreign standards without consideration of geologic setting and lack of independent verification steps. Therefore, we suggest a carefully designed verification studies from a test site. We also suggest that care must be exercised when applying construction standards for the interpretation and accessment of rock mass properties.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.9 no.6
• /
• pp.732-742
• /
• 1985

Measurements of mean velocity and turbulence characteristics are obtained with a linearized constant temperature hot-wire anemometer in a two-dimensional turbulent jet discharging parallel to a flate. Wall static pressure distribution is also measure. The Reynolds number based on the jet nozzle width (D) is about 42,000 and the step height is 2.5D. The reattachment length is found to be 7.5D by using both wool tuft and oil methods. Upstream of the reattachment point, there exist double coherent structures and mean velocity, Reynolds stresses and triple product profiles are asymmetric about jet center line due to the influence of streamline curvature and recirculating flow region. Near the reattachment point, wall static pressure and turbulence quantities change its shape rapidly because of the large eddies by the solid wall. Especially, turbulence intensity has a maximum value in the reattachment regin, then decreases slowly in the redeveloping wall jet ragion. Downstream of X/D=14, a single large scale eddy structure is formed. Far downstream affer the reattachment(X/D.geq.18) mean velocity profile, the decay of maximum velocity and the variation of jet half width are nearly similar to those of plane wall jet, but the Reynolds stresses are higher than those of the latter.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.6D
• /
• pp.685-691
• /
• 2009

The purpose of this paper is to classify the traffic condition in freeway by the impacts of road weather. The factor analysis showed that weather factors, which are considered as influential, are identified as weather condition (rain or clear), temperature and sight distance with RWIS and VDS data in Seohae bridge used. The result of ANOVA shows that weather is dividedinto clear and rainy; temperature into below and equal or above $5^{\circ}C$ and sight distance into below or equal or above 10km. Based on those factors, the freeway traffic condition has been classified as five different types. The flow-speed model for each traffic conditions was proposed, which was not significant due to the lack of smaple data. Although not sufficient, the methodology to categorize traffic situation model presented in this paper may shed light on the idea for the future and can be used for proper traffic management for each weather condition.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.5
• /
• pp.483-488
• /
• 2015

This paper presents a local binary pattern(LBP) for effectively classifying textures, which is based on the multidimensional intensity difference between the adjacent pixels in the block image. The intensity difference by considering the a extent of 4 directional changes(verticality, horizontality, diagonality, inverse diagonality) in brightness between the adjacent pixels is applied to reduce the computation load as a results of decreasing the levels of histogram for classifying textures of image. And the binary patterns that is represented by the relevant intensities within a block image, is also used to effectively classify the textures by accurately reflecting the local attributes. The proposed method has been applied to classify 24 block images from USC Texture Mosaic #2 of 128*128 pixels gray image. The block images are different in size and texture. The experimental results show that the proposed method has a speedy classification and makes a free size block images classify possible. In particular, the proposed method gives better results than the conventional LBP by increasing the range of histogram level reduction as the block size becomes larger.

• Investigative Magnetic Resonance Imaging
• /
• v.6 no.2
• /
• pp.137-146
• /
• 2002

Purpose : This paper introduces an improved classification approach which adopts a deterministic relaxation method and an agglomerative clustering technique for the classification of MRI using neural network. The proposed approach can solve the problems of convergency to local optima and computational burden caused by a large number of input patterns when a neural network is used for image classification. Materials and methods : Application of Hopfield neural network has been solving various optimization problems. However, major problem of mapping an image classification problem into a neural network is that network is opt to converge to local optima and its convergency toward the global solution with a standard stochastic relaxation spends much time. Therefore, to avoid local solutions and to achieve fast convergency toward a global optimization, we adopt MFA to a Hopfield network during the classification. MFA replaces the stochastic nature of simulated annealing method with a set of deterministic update rules that act on the average value of the variable. By minimizing averages, it is possible to converge to an equilibrium state considerably faster than standard simulated annealing method. Moreover, the proposed agglomerative clustering algorithm which determines the underlying clusters of the image provides initial input values of Hopfield neural network. Results : The proposed approach which uses agglomerative clustering and deterministic relaxation approach resolves the problem of local optimization and achieves fast convergency toward a global optimization when a neural network is used for MRI classification. Conclusion : In this paper, we introduce a new paradigm to classify MRI using clustering analysis and deterministic relaxation for neural network to improve the classification results.

• The KIPS Transactions:PartB
• /
• v.8B no.4
• /
• pp.335-342
• /
• 2001

본 논문은 보로노이 공간분류를 활용하여 원격탐사 영상인식을 위한 다층 신경망 분류기를제안한다. 제안된 다층 신경망 분류기는 보로노이 다각형 영역으로 클래스를 구분하며, 초평면 방정식의 계수를 오류 역전과 학습 초기의 연결 강도, 임계치 그리고 은닉층의 노드 수로 결정한다. 제안된 방법은 오류역전과 학습 알고리즘에서 임의로 정해주던 초기 정보를 사전 분석에 의해 공학적으로 결정함으로써 느린 수렴 속도와 학습실패 등의 단점을 피할 수 있는 장점이 있다. 보로노이 다이어그램에 대한 경계선의 초평면 방정식은 훈련집합의 클래스별 평균값을 구하여 Mathematica 패키지로 계산하였다. 제안된 다층 신경망에 의한 영상분류기의 인식능력을 평가하기 위하여 원격탐사 영상인식에서 자주 활용되는 최소거리 분류 방법과 최대우도 분류 방법으로 처리해서 비교한 결과, 최소거리 분류 방법은 실험화상에 대해 81.4%, 최대우도 부류기에 의한 분류는 87.8%, 제안한 방법은 92.2% 정확성을 가진 분류결과를 나타냈다.