Inference and parameter estimation on belief networks for image segmentation
نویسندگان
چکیده
We introduce a new causal hierarchical belief network for image segmentation. Contrary to classical tree structured (or pyramidal) models, the factor graph of the network contains cycles. Each level of the hierarchical structure features the same number of sites as the base level and each site on a given level has several neighbors on the parent level. Compared to tree structured models, the (spatial) random process on the base level of the model is stationary which avoids known drawbacks, namely visual artifacts in the segmented image. We propose different parametrisations of the conditional probability distributions governing the transitions between the image levels. A parametric distribution depending on a single parameter allows the design of a fast inference algorithm on graph cuts, whereas the parameter is estimated with a least squares technique. For arbitrary distributions, we propose inference with loopy belief propagation and we introduce a new parameter estimation technique adapted to the model. RÉSUMÉ. Dans ce papier, on présente un nouveau réseau bayesien hiérarchique dédié à la segmentation d’images. Contrairement aux modèles classiques (e.g. le quad arbre), le graphe associé à ce réseau bayesien contient des cycles. Chaque niveau de cette structure hiérarchique contient autant de sommets que le niveau de base (qui contient lui-même autant de sommets que l’image à segmenter contient de pixels) et chaque sommet a plusieurs parents sur le niveau supérieur. Contrairement aux structures d’arbre classiques, ce modèle possède la propriété de stationnarité ce qui signifie qu’il est invariant aux translations de l’image. Le présence de boucles dans le modèle proposé rend le problème de l’inférence exacte délicat. L’apport majeur de ce papier, outre le modèle lui-même, est la présentation d’un algorithme d’inférence exact basé sur le problème classique de graphes, à base de maximisation de flots.
منابع مشابه
Inference and parameter estimation on hierarchical belief networks for image segmentation
We introduce a new causal hierarchical belief network for image segmentation. Contrary to classical tree structured (or pyramidal) models, the factor graph of the network contains cycles. Each level of the hierarchical structure features the same number of sites as the base level and each site on a given level has several neighbors on the parent level. Compared to tree structured models, the (s...
متن کاملA Surface Water Evaporation Estimation Model Using Bayesian Belief Networks with an Application to the Persian Gulf
Evaporation phenomena is a effective climate component on water resources management and has special importance in agriculture. In this paper, Bayesian belief networks (BBNs) as a non-linear modeling technique provide an evaporation estimation method under uncertainty. As a case study, we estimated the surface water evaporation of the Persian Gulf and worked with a dataset of observations ...
متن کاملA Surface Water Evaporation Estimation Model Using Bayesian Belief Networks with an Application to the Persian Gulf
Evaporation phenomena is a effective climate component on water resources management and has special importance in agriculture. In this paper, Bayesian belief networks (BBNs) as a non-linear modeling technique provide an evaporation estimation method under uncertainty. As a case study, we estimated the surface water evaporation of the Persian Gulf and worked with a dataset of observations ...
متن کاملMotion Segmentation Using Inference in Dynamic Bayesian Networks
Existing formulations for optical flow estimation and image segmentation have used Bayesian Networks and Markov Random Field (MRF) priors to impose smoothness of segmentation. These approaches typically focus on estimation in a single time slice based on two consecutive images. We develop a motion segmentation framework for a continuous stream of images using inference in a corresponding Dynami...
متن کاملRobust Potato Color Image Segmentation using Adaptive Fuzzy Inference System
Potato image segmentation is an important part of image-based potato defect detection. This paper presents a robust potato color image segmentation through a combination of a fuzzy rule based system, an image thresholding based on Genetic Algorithm (GA) optimization and morphological operators. The proposed potato color image segmentation is robust against variation of background, distance and ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2008