Iterative Local-Global Energy Minimization for Automatic Extraction of Objects of Interest
IEEE Transactions on Pattern Analysis and Machine Intelligence
Robust Image Segmentation Using Resampling and Shape Constraints
IEEE Transactions on Pattern Analysis and Machine Intelligence
Histogram based segmentation using Wasserstein distances
SSVM'07 Proceedings of the 1st international conference on Scale space and variational methods in computer vision
Tensor-SIFT Based Earth Mover's Distance for Contour Tracking
Journal of Mathematical Imaging and Vision
Active contour model driven by local histogram fitting energy
Pattern Recognition Letters
| Hi-index | 0.00 |
We consider the problem of image segmentation using active contoursthrough the minimization of an energy criterion involving bothregion and boundary functionals. These functionals are derivedthrough a shape derivative approach instead of classical calculusof variation. The equations can be elegantly derived withoutconverting the region integrals into boundary integrals. From thederivative, we deduce the evolution equation of an active contourthat makes it evolve towards a minimum of the criterion. We focusmore particularly on statistical features globally attached to theregion and especially to the probability density functions of imagefeatures such as the color histogram of a region. A theoreticalframework is set for the minimization of the distance between twohistograms for matching or tracking purposes. An application ofthis framework to the segmentation of color histograms in videosequences is then proposed. We briefly describe our numericalscheme and show some experimental results.