Variational segmentation of image sequences using deformable shape priors

Authors:
Ketut Fundana;Niels Chr. Overgaard;Anders Heyden
Affiliations:
Applied Mathematics Group, School of Technology and Society, Malmö University, Sweden;Applied Mathematics Group, School of Technology and Society, Malmö University, Sweden;Applied Mathematics Group, School of Technology and Society, Malmö University, Sweden
Venue:
SCIA'07 Proceedings of the 15th Scandinavian conference on Image analysis
Year:
2007

Citing 8
Cited 1

Geodesic Active Contours

International Journal of Computer Vision
Geodesic Active Contours and Level Sets for the Detection and Tracking of Moving Objects

IEEE Transactions on Pattern Analysis and Machine Intelligence
Shape Priors for Level Set Representations

ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part II
Geodesic active regions and level set methods for motion estimation and tracking

Computer Vision and Image Understanding
Towards recognition-based variational segmentation using shape priors and dynamic labeling

Scale Space'03 Proceedings of the 4th international conference on Scale space methods in computer vision
A geometric formulation of gradient descent for variational problems with moving surfaces

Scale-Space'05 Proceedings of the 5th international conference on Scale Space and PDE Methods in Computer Vision
Dynamical statistical shape priors for level set based sequence segmentation

VLSM'05 Proceedings of the Third international conference on Variational, Geometric, and Level Set Methods in Computer Vision
Active contours without edges

IEEE Transactions on Image Processing

Variational Segmentation of Image Sequences Using Region-Based Active Contours and Deformable Shape Priors

International Journal of Computer Vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

The segmentation of objects in image sequences is an important and difficult problem in computer vision with applications to e.g. video surveillance. In this paper we propose a new method for variational segmentation of image sequences containing nonrigid, moving objects. The method is based on the classical Chan-Vese model augmented with a novel frame-to-frame interaction term, which allow us to update the segmentation result from one image frame to the next using the previous segmentation result as a shape prior. The interaction term is constructed to be pose-invariant and to allow moderate deformations in shape. It is expected to handle the appearance of occlusions which otherwise can make segmentation fail. The performance of the model is illustrated with experiments on real image sequences.