Globally optimal joint image segmentation and shape matching based on Wasserstein modes
From MaRDI portal
Publication:890098
DOI10.1007/S10851-014-0546-8zbMATH Open1343.68277arXiv1407.3956OpenAlexW2019588894MaRDI QIDQ890098FDOQ890098
Authors: Bernhard Schmitzer, Christoph Schnörr 
Publication date: 9 November 2015
Published in: Journal of Mathematical Imaging and Vision (Search for Journal in Brave)
Abstract: A functional for joint variational object segmentation and shape matching is developed. The formulation is based on optimal transport w.r.t. geometric distance and local feature similarity. Geometric invariance and modelling of object-typical statistical variations is achieved by introducing degrees of freedom that describe transformations and deformations of the shape template. The shape model is mathematically equivalent to contour-based approaches but inference can be performed without conversion between the contour and region representations, allowing combination with other convex segmentation approaches and simplifying optimization. While the overall functional is non-convex, non-convexity is confined to a low-dimensional variable. We propose a locally optimal alternating optimization scheme and a globally optimal branch and bound scheme, based on adaptive convex relaxation. Combining both methods allows to eliminate the delicate initialization problem inherent to many contour based approaches while remaining computationally practical. The properties of the functional, its ability to adapt to a wide range of input data structures and the different optimization schemes are illustrated and compared by numerical experiments.
Full work available at URL: https://arxiv.org/abs/1407.3956
Recommendations
- Modelling convex shape priors and matching based on the Gromov-Wasserstein distance
- Convex histogram-based joint image segmentation with regularized optimal transport cost
- scientific article; zbMATH DE number 2062591
- Nonlocal joint segmentation registration model
- Sparse template-based variational image segmentation
Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10)
Cites Work
- Riemannian geometries on spaces of plane curves
- Computing large deformation metric mappings via geodesic flows of diffeomorphisms
- Shapes and diffeomorphisms
- A metric on shape space with explicit geodesics
- Barycenters in the Wasserstein space
- Title not available (Why is that?)
- Optimal Transport
- Nonlocal Operators with Applications to Image Processing
- Algorithms for Finding Global Minimizers of Image Segmentation and Denoising Models
- Title not available (Why is that?)
- A new geometric metric in the space of curves, and applications to tracking deforming objects by prediction and filtering
- On shape of plane elastic curves
- Optimal mass transport for registration and warping
- Some geometric calculations on Wasserstein space
- Continuous multiclass labeling approaches and algorithms
- Efficient domination in cubic vertex-transitive graphs
- A linear optimal transportation framework for quantifying and visualizing variations in sets of images
- Approximations of shape metrics and application to shape warping and empirical shape statis\-tics
- Gromov-Wasserstein distances and the metric approach to object matching
- Shape statistics in kernel space for variational image segmentation.
- Branch-and-mincut: global optimization for image segmentation with high-level priors
- Modelling convex shape priors and matching based on the Gromov-Wasserstein distance
- Efficient Computation of Isometry‐Invariant Distances Between Surfaces
- Functional currents: a new mathematical tool to model and analyse functional shapes
Cited In (7)
- Matrix Balancing Based Interior Point Methods for Point Set Matching Problems
- Least action principles for incompressible flows and geodesics between shapes
- A sparse multiscale algorithm for dense optimal transport
- Modelling convex shape priors and matching based on the Gromov-Wasserstein distance
- Convex histogram-based joint image segmentation with regularized optimal transport cost
- Scott: shape-location combined tracking with optimal transport
- Stabilized Sparse Scaling Algorithms for Entropy Regularized Transport Problems
Uses Software
This page was built for publication: Globally optimal joint image segmentation and shape matching based on Wasserstein modes
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q890098)
