• 제목/요약/키워드: diffusion probabilistic model

검색결과 14건 처리시간 0.024초

확률적 확산을 이용한 스테레오 정합 알고리듬 (New stereo matching algorithm based on probabilistic diffusion)

  • 이상화;이충웅
    • 전자공학회논문지S
    • /
    • 제35S권4호
    • /
    • pp.105-117
    • /
    • 1998
  • In this paper, the general formula of disparity estimation based on Bayesian Maximum A Posteriori (MAP) algorithm is derived and implemented with simplified probabilistic models. The probabilistic models are independence and similarity among the neighboring disparities in the configuration.The formula is the generalized probabilistic diffusion equation based on Bayesian model, and can be implemented into the some different forms corresponding to the probabilistic models in the disparity neighborhood system or configuration. And, we proposed new probabilistic models in order to simplify the joint probability distribution of disparities in the configuration. According to the experimental results, the proposed algorithm outperformed the other ones, such as sum of swuared difference(SSD) based algorithm and Scharstein's method. We canconclude that the derived formular generalizes the probabilistic diffusion based on Bayesian MAP algorithm for disparity estimation, and the propsoed probabilistic models are reasonable and approximate the pure joint probability distribution very well with decreasing the computations to 0.01% of the generalized formula.

  • PDF

Probabilistic time-dependent sensitivity analysis of HPC bridge deck exposed to chlorides

  • Ghosh, Pratanu;Konecny, Petr;Lehner, Petr;Tikalsky, Paul J.
    • Computers and Concrete
    • /
    • 제19권3호
    • /
    • pp.305-313
    • /
    • 2017
  • A robust finite element based reinforced concrete bridge deck corrosion initiation model is applied for time-dependent probabilistic sensitivity analysis. The model is focused on uncertainties in the governing parameters that include variation of high performance concrete (HPC) diffusion coefficients, concrete cover depth, surface chloride concentration, holidays in reinforcements, coatings and critical chloride threshold level in several steel reinforcements. The corrosion initiation risk is expressed in the form of probability over intended life span of the bridge deck. Conducted study shows the time-dependent sensitivity analysis to evaluate the significance of governing parameters on chloride ingress rate, various steel reinforcement protection and the corrosion initiation likelihood. Results from this probabilistic analysis provide better insight into the effect of input parameters variation on the estimate of the corrosion initiation risk for the design of concrete structures in harsh chloride environments.

콘크리트 구조물의 확률론적 탄산화 예측 모델 개발 및 내구성 해석 (Durability Analysis and Development of Probability-Based Carbonation Prediction Model in Concrete Structure)

  • 정현준
    • 대한토목학회논문집
    • /
    • 제30권4A호
    • /
    • pp.343-352
    • /
    • 2010
  • 최근에 탄산화 콘크리트 구조물의 정량적인 사용수명과 장기적인 성능을 확보하고 예측하기 위해서 확률론적인 내구성 해석 및 설계를 수행하는 연구가 많이 진행되고 있다. 이와 관련하여 콘크리트 구조물에 확률론적 내구성 설계 개념을 도입되고 있다. 본 논문에서는 탄산화 콘크리트 구조물의 통계적인 자료를 이용하여 Fick의 첫 번째 법칙에 근거한 탄산화 예측 모델에 적용하였으며, 이를 이용하여 확률론적 내구성 해석을 수행하였다. 이 예측모델에 관련된 설계변수인 $CO_2$ 확산계수, 대기중의 $CO_2$ 농도, $CO_2$ 흡착량, 시멘트 수화도 등의 영향을 검토하였다. 확률론에 기초한 탄산화 예측모델은 여러 환경에 위치한 콘크리트 구조물에 모니터링 자료를 이용하여 탄산화 깊이와 잔존수명을 예측하였다. 그 결과로 본 연구에서 합리적인 탄산화 예측모델을 이용한 적용 방법은 탄산화 콘크리트 구조물의 내구성 확보 및 구조물의 손상 개시시기를 예측하고 구조물을 유지 관리하기 위한 유연한 의사결정을 할 수 있을 것으로 판단된다.

Probabilistic evaluation of chloride ingress process in concrete structures considering environmental characteristics

  • Taisen, Zhao;Yi, Zhang;Kefei, Li;Junjie, Wang
    • Structural Engineering and Mechanics
    • /
    • 제84권6호
    • /
    • pp.831-849
    • /
    • 2022
  • One of the most prevalent causes of reinforced concrete (RC) structural deterioration is chloride-induced corrosion. This paper aims to provide a comprehensive insight into the environmental effect of RC's chloride ingress process. The first step is to investigate how relative humidity, temperature, and wind influence chloride ingress into concrete. The probability of initiation time of chloride-induced corrosion is predicted using a probabilistic model that considers these aspects. Parametric analysis is conducted on several factors impacting the corrosion process, including the depth of concrete cover, surface chloride concentration, relative humidity, and temperature to expose environmental features. According to the findings, environmental factors such as surface chloride concentration, relative humidity and temperature substantially impact on the time to corrosion initiation. The long- and short-distance impacts are also examined. The meteorological data from the National Meteorological Center of China are collected and used to analyze the environmental characteristics of the chloride ingress issue for structures along China's coastline. Finally, various recommendations are made for improving durability design against chloride attacks.

GRF기반이방성 분산 복원에 의한 분류 결과 향상 (Improvement in Image Classification by GRF-based Anisotropic Diffusion Restoration)

  • 이상훈
    • 한국GIS학회:학술대회논문집
    • /
    • 한국GIS학회 2004년도 GIS/RS 공동 춘계학술대회 논문집
    • /
    • pp.523-528
    • /
    • 2004
  • 본 연구는 원격탐사 영상분류 과정에 이방성 분산 복원의 적용을 제안하고 있다. 수정 이방성 분산 복원은 많은 원격탐사 영상에 나타나는 지리적 연결성을 대표하는 Markov random field에 기반한 확률적 모형을 사용하고 있고 반복적인 확산과정을 통해 영상복원을 수행한다. 제안 확산과정은 지리적 연결성과 연관된 응집력 계수를 위하여 brightness gradient의 함수를 사용하며 매 반복단계마다 adaptive하게 추정한다. 한반도의 위성 원격탐사 자료에 대한 실험을 실시하였고 제안된 수정 이방성 분산 복원의 적용은 실제 관측 자료에도 매우 효과적임을 알 수 있었다.

  • PDF

수정 이방성 분산 복원을 이용한 영상 분류 (Image Classification Using Modified Anisotropic Diffusion Restoration)

  • 이상훈
    • 대한원격탐사학회지
    • /
    • 제19권6호
    • /
    • pp.479-490
    • /
    • 2003
  • This study proposed a modified anisotropic diffusion restoration for image classification. The anisotropic diffusion restoration uses a probabilistic model based on Markov random field, which represents geographical connectedness existing in many remotely sensed images, and restores them through an iterative diffusion processing. In every iteration, the bonding-strength coefficient associated with the spatial connectedness is adaptively estimated as a function of brightness gradient. The gradient function involves a constant called "temperature", which determines the amount of discontinuity and is continuously decreased in the iterations. In this study, the proposed method has been extensively evaluated using simulated images that were generated from various patterns. These patterns represent the types of natural and artificial land-use. The simulated images were restored by the modified anisotropic diffusion technique, and then classified by a multistage hierarchical clustering classification. The classification results were compared to them of the non-restored simulation images. The restoration with an appropriate temperature considerably reduces error in classification, especially for noisy images. This study made experiments on the satellite images remotely sensed on the Korean peninsula. The experimental results show that the proposed approach is also very effective on image classification in remote sensing.

전력수요측의 고효율기기 확산을 고려한 전원개발계획 비교실험 (A Comparative Simulation on Generation Expansion Planning Considering the Diffusion of High Efficient End-Uses)

  • 장승찬;조형준;김발호;김정훈
    • 대한전기학회:학술대회논문집
    • /
    • 대한전기학회 2000년도 하계학술대회 논문집 A
    • /
    • pp.328-330
    • /
    • 2000
  • This paper presents an experimental comparison between the conventional generation expansion planning and DSM incorporated one. As a DSM measure, diffusion of high efficient end-uses is considered and its impact will be targeted at the strategic energy conservation. A revised probabilistic production simulation is proposed by modifying the equivalent load duration curve (ELDC) with the capacity deconvolution of DSM end-use. To investigate long-term DSM impacts relative to the conventional planning, WASP model is applied and the effectiveness of DSM planning as an electricity resources is demonstrated.

  • PDF

Prediction of RC structure service life from field long term chloride diffusion

  • Safehian, Majid;Ramezanianpour, Ali Akbar
    • Computers and Concrete
    • /
    • 제15권4호
    • /
    • pp.589-606
    • /
    • 2015
  • It is well-documented that the major deterioration of coastal RC structures is chloride-induced corrosion. Therefore, regional investigations are necessary for durability based design and evaluation of the proposed service life prdiction models. In this paper, four reinforced concrete jetties exposed to severe marine environment were monitored to assess the long term chloride penetration at 6 months to 96 months. Also, some accelerated durability tests were performed on standard samples in laboratory. As a result, two time-dependent equations are proposed for basic parameters of chloride diffusion into concrete and then the corrosion initiation time is estimated by a developed probabilistic service life model Also, two famous service life prediction models are compared using chloride profiles obtained from structures after about 40 years in the tidal exposure conditions. The results confirm that the influence of concrete quality on diffusion coefficients is related to the concrete pore structure and the time dependence is due to chemical reactions of sea water ions with hydration products which lead a reduction in pore structure. Also, proper attention to the durability properties of concrete may extend the service life of marine structures greater than fifty years, even in harsh environments.

디퓨전 오토인코더의 시선 조작 데이터 증강을 통한 시선 추적 (Gaze-Manipulated Data Augmentation for Gaze Estimation With Diffusion Autoencoders)

  • 문강륜;김영한;박용준;김용규
    • 한국컴퓨터그래픽스학회논문지
    • /
    • 제30권3호
    • /
    • pp.51-59
    • /
    • 2024
  • 시선 벡터 정답값을 갖는 대규모 데이터의 수집은 시선 추적 분야에서 많은 비용을 필요로 한다. 본 논문에서는 원본 사진의 시선을 수정하는 데이터 증강 기법을 사용하여 제한된 개수의 시선 정답값이 주어진 상황에서 시선 추적 모델의 정확도를 향상시키는 방법을 제안한다. 시선 구간 다중 클래스 분류를 보조 작업으로 학습하고, 디퓨전 오토인코더의 잠재 변수를 조정하여 원본 사진의 시선을 편집한 사진을 생성한다. 기존의 얼굴 속성 편집과 달리, 우리는 이진 속성이 아닌 시선 벡터의 피치와 요를 지정한 범주 내로 변경하며, 편집된 사진을 시선 추적 모델의 증강된 학습 데이터로 활용한다. 시선 정답값이 5만 개 이하일 때 준지도 학습에서의 시선 추적 모델의 정확도 향상은 제안한 데이터 증강 기법의 효과를 입증한다.

Wind-induced random vibration of saddle membrane structures: Theoretical and experimental study

  • Rongjie Pan;Changjiang Liu;Dong Li;Yuanjun Sun;Weibin Huang;Ziye Chen
    • Wind and Structures
    • /
    • 제36권2호
    • /
    • pp.133-147
    • /
    • 2023
  • The random vibration of saddle membrane structures under wind load is studied theoretically and experimentally. First, the nonlinear random vibration differential equations of saddle membrane structures under wind loads are established based on von Karman's large deflection theory, thin shell theory and potential flow theory. The probabilistic density function (PDF) and its corresponding statistical parameters of the displacement response of membrane structure are obtained by using the diffusion process theory and the Fokker Planck Kolmogorov equation method (FPK) to solve the equation. Furthermore, a wind tunnel test is carried out to obtain the displacement time history data of the test model under wind load, and the statistical characteristics of the displacement time history of the prototype model are obtained by similarity theory and probability statistics method. Finally, the rationality of the theoretical model is verified by comparing the experimental model with the theoretical model. The results show that the theoretical model agrees with the experimental model, and the random vibration response can be effectively reduced by increasing the initial pretension force and the rise-span ratio within a certain range. The research methods can provide a theoretical reference for the random vibration of the membrane structure, and also be the foundation of structural reliability of membrane structure based on wind-induced response.