Image segmentation using clique based shape prior and the Mumford Shah Functional

Park, Fredrick

doi:10.1109/icip.2015.7351572

Cited by 1 publication

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following the work of Park [34], we propose to incorporate a two dimensional variant of the clique based signature proposed by Kimmel et al [3,4,15] into the GAC model. In certain settings, the edge based segmentation approach can have advantages over a region based one like that found in the Mumford-Shah Model (MS) [30], which in its 2-phase piecewise constant formulation is known as the Chan and Vese model (CV) [8].…”

Section: Introductionmentioning

confidence: 99%

“…The closest related work is that of Park [34] where the author utilizes a polygonal implementation of the 2-phase MS functional and a cliques shape prior signature. The author shows successful numerical results for segmenting occluded objects and objects with nonuniform intensities in the presence of clutter.…”

Section: Introductionmentioning

confidence: 99%

“…The author shows successful numerical results for segmenting occluded objects and objects with nonuniform intensities in the presence of clutter. Despite using the same shape signature, the proposed model differs significantly from that in [34] by using the GAC energy instead of the MS energy. A key benefit in the proposed approach is the ability to segment objects that rely more on features instead of object contrast.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Geodesic Active Contours with Shape Priors for Segmentation, Disocclusion, and Illusory Contour Capture

Householder¹

2020

SIURO

View full text Add to dashboard Cite

Image segmentation is the task of finding salient regions of importance in an image. In this work, we take a curve evolution approach to this problem where we deform an initial curve in the inward normal direction with the objective of finding the boundaries of objects in an image. To achieve this, we propose a variational image segmentation model that incorporates a clique based shape signature with a geodesic active contours energy. The model scheme consists of evolving a parametric representation of an active contour to minimize the penalty that the model induces. This penalty is minimized when the curve is on the boundaries of objects in the image, areas with sudden change in pixel intensity i.e. light to dark. We demonstrate successful capture of illusory contours, segmentation of objects in a cluttered background, and segmentation of occluded objects.

show abstract