Level set based segmentation with intensity and curvature priors

Leventon, Michael E.; Faugeras, Olivier; Grimson, W. Eric L.; Well, W.M.

doi:10.1109/ssbi.2002.1233988

Cited by 19 publications

(18 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Landmark-based priors, also called active shape models [110][111][112][113] have been used in an extensive range of applications. More recently, priors have been used in spaces of curves, either via a model centered on level-set representations [114][115][116][117][118][119][120][121][122][123] or directly, notably using shape distances to a template [124,125]. The standard Bayesian approach assumes that a prior distribution, say P 0 , has been defined over a shape space M, together with an observation scheme formalized as a conditional distribution of observables (discrete curves, images, …) given the unknown random shape, that can be denoted P(observ|c).…”

Section: Applications: Analyzing Shape Space Datamentioning

confidence: 99%

Spaces and manifolds of shapes in computer vision: An overview

Younes

2012

Image and Vision Computing

View full text Add to dashboard Cite

Section: Applications: Analyzing Shape Space Datamentioning

confidence: 99%

Spaces and manifolds of shapes in computer vision: An overview

Younes

2012

Image and Vision Computing

View full text Add to dashboard Cite

“…Level set method was firstly introduced by Osher and Sethian [5], and is now widely used in image processing such as image segmentation, registration, and detection of moving objects [6][7][8] of higher-dimensional function φ . The evolution equation of the level set function φ can be written in the following general form:…”

Section: A Level Set Theorymentioning

confidence: 99%

A multistage processing procedure for 4D breast MRI segmentation

Wang

Ding

Wang

2008

2008 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society

View full text Add to dashboard Cite

Contrast-enhanced magnetic resonance imaging (MRI) is a novel approach to detect breast tumor, however, multiple 3D image sets need to be analyzed which causes unaffordable inspection tasks to physicians. Computer assistant detection is of great help to this situation and image segmentation is most important process of computer assistant detection. To segment the breast region from 3D MRI set, a multistage image processing procedure was proposed. Considering the morphological features of breast image, the image was divided into two parts to reduce the complexity of the segmentation procedure. For the anterior part of the image, the breast tissue was segmented from background using the region growing method. Meanwhile, the mean intensity of breast tissue was estimated. To segment the posterior part of the image, the estimated mean breast intensity was used to initialize an approximate boundary between breast and chest. The estimated intensity distribution of breast tissue was used to set the propagation term of following level set iterations. Based on this initialization, threshold-based 3D level set algorithm was used to seek the precise boundary between chest and breast. The level set algorithm achieves novel performance for the 3D segmentation of chest boundary. Based on the segmentation of one set of 3D image, the segmentation result was used as the initial boundary of following image sets to achieve automatic 4D segmentation. This multistage procedure greatly accelerates the convergence of the level set algorithm and reduces the chance of running into local minimum. Clinical data demonstrated that this processing procedure was effective for automatic segmentation of 4D breast MRI.

show abstract

“…Local geometric properties (curvature, local smoothness constraints) can be used when defining such prior model (Leventon et al 2000a) or it can be done in a more global manner (Cremers et al 2001) that captures the variance of the entire structure of interest. While local models are quite efficient, global representations are more appropriate to cope with occlusions, noise, and changes on the object pose.…”

Section: Introductionmentioning

confidence: 99%

Prior Knowledge, Level Set Representations & Visual Grouping

Rousson

Paragios

2007

Int J Comput Vis

View full text Add to dashboard Cite

International audienceIn this paper, we propose a level set method for shape-driven object extraction. We introduce a voxel-wise probabilistic level set formulation to account for prior knowledge. To this end, objects are represented in an implicit form. Constraints on the segmentation process are imposed by seeking a projection to the image plane of the prior model modulo a similarity transformation. The optimization of a statistical metric between the evolving contour and the model leads to motion equations that evolve the contour toward the desired image properties while recovering the pose of the object in the new image. Upon convergence, a solution that is similarity invariant with respect to the model and the corresponding transformation are recovered. Promising experimental results demonstrate the potential of such an approach

show abstract

Level set based segmentation with intensity and curvature priors

Cited by 19 publications

References 18 publications

Spaces and manifolds of shapes in computer vision: An overview

Spaces and manifolds of shapes in computer vision: An overview

A multistage processing procedure for 4D breast MRI segmentation

Prior Knowledge, Level Set Representations & Visual Grouping

Contact Info

Product

Resources

About