Level Lines as Global Minimizers of Energy Functionals in Image Segmentation

Kervrann, Charles; Hoebeke, Mark; Trubuil, Alain

doi:10.1007/3-540-45053-x_16

Cited by 8 publications

(11 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…related to the Markov connected components fields, has been already discussed by Kervrann et al (2000), Alvarez et al (1999) and Møller and Waagepetersen (1998).…”

Section: The Bayesian Frameworkmentioning

confidence: 86%

“…The regularization parameter λ can be then interpreted as a scale parameter that only tunes the number of regions (Morel and Solimini, 1994;Kervrann et al, 2000). If λ § 0, each point is potentially a region and Ω § / 0 ; the global minimum coincides with zero and this segmentation is called the "trivial segmentation" (Morel and Solimini, 1994).…”

Section: Bayesian Inferencementioning

confidence: 99%

See 1 more Smart Citation

Bayesian Image Segmentation Through Level Lines Selection

Kervrann¹

2011

Image Anal Stereol

Self Cite

View full text Add to dashboard Cite

Bayesian statistical theory is a convenient way of taking a priori information into consideration when inference is made from images. In Bayesian image segmentation, the a priori distribution should capture the knowledge about objects. Taking inspiration from (Alvarez et al., 1999), we design a prior density that penalizes the area of homogeneous parts in images. The segmentation problem is further formulated as the estimation of the set of curves that maximizes the posterior distribution. In this paper, we explore a posterior distribution model for which its maximal mode is given by a subset of level curves, that is the boundaries of image level sets. For the completeness of the paper, we present a stepwise greedy algorithm for computing partitions with connected components.

show abstract

“…related to the Markov connected components fields, has been already discussed by Kervrann et al (2000), Alvarez et al (1999) and Møller and Waagepetersen (1998).…”

Section: The Bayesian Frameworkmentioning

confidence: 86%

Section: Bayesian Inferencementioning

confidence: 99%

Bayesian Image Segmentation Through Level Lines Selection

Kervrann¹

2011

Image Anal Stereol

Self Cite

View full text Add to dashboard Cite

show abstract

“…Hence, a segmentation of the MIP map can be used to detect the OD regions in the image. By applying the segmentation method described in [12] to the MIP map of Fig. 2 (middle), meaningful regions are extracted as shown in Fig.…”

Section: Extraction Of Origin/destination Regionsmentioning

confidence: 99%

Minimal paths and probabilistic models for origin-destination traffic estimation in live cell imaging

Pécot

Kervrann

Bouthémy

2008

2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro

View full text Add to dashboard Cite

Green Fluorescent Protein (GFP)-tagging and time-lapse fluorescence microscopy enable to observe molecular dynamics and interactions in live cells. Original image analysis methods are then required to process challenging 2D or 3D image sequences. To address the tracking problem of several hundreds of objects, we propose an original framework that provides general information about vesicle transport, that is traffic flows between origin and destination regions detected in the image sequence. Traffic estimation can be accomplished by adapting the advances in Network Tomography commonly used in network communications. In this paper, we address image partition given vesicle stocking areas and multipaths routing for vesicle transport. This approach has been developed for real fluorescence image sequences and Rab proteins.

show abstract

“…Additionally, the variables {|Ω i |} may be considered as independent random variables with density g(|Ω i |). Hence, the prior distribution is of the form π( γ has been already discussed in [13,1]. This model is related to the Markov connected components fields [16].…”

Section: The Bayesian Frameworkmentioning

confidence: 99%

“…The penalty functional tends to regulate the emergence of objects Ω i in the image and gives no control on the smoothness of boundaries. The regularization parameter λ can be then interpreted as a scale parameter that only tunes the number of regions [17,14,13]. If λ = 0, each point is potentially a region and Ω = ∅ ; the global minimum coincides with zero and this segmentation is called the "trivial segmentation" [17,14].…”

Section: Bayesian Inferencementioning

confidence: 99%

Bayesian Object Detection through Level Curves Selection

Kervrann

2001

Scale-Space and Morphology in Computer Vision

Self Cite

View full text Add to dashboard Cite

Abstract. Bayesian statistical theory is a convenient way of taking a priori information into consideration when inference is made from images. In Bayesian image detection, the a priori distribution should capture the knowledge about objects. Taking inspiration from [1], we design a prior density that penalizes the area of homogeneous parts in images. The detection problem is further formulated as the estimation of the set of curves that maximizes the posterior distribution. In this paper, we explore a posterior distribution model for which its maximal mode is given by a subset of level curves, that is the boundaries of image level sets. For the completeness of the paper, we present a stepwise greedy algorithm for computing partitions with connected components.

show abstract

Level Lines as Global Minimizers of Energy Functionals in Image Segmentation

Cited by 8 publications

References 23 publications

Bayesian Image Segmentation Through Level Lines Selection

Bayesian Image Segmentation Through Level Lines Selection

Minimal paths and probabilistic models for origin-destination traffic estimation in live cell imaging

Bayesian Object Detection through Level Curves Selection

Contact Info

Product

Resources

About