Mean shift-based Bayesian image reconstruction into visual subspace.
Torbjørn VikFabrice HeitzPierre CharbonnierPublished in: ICIP (1) (2003)
Keyphrases
- image reconstruction
- mean shift
- feature space
- reconstructed image
- back projection
- object tracking
- graph cuts
- super resolution
- compressed sensing
- mode seeking
- reconstruction method
- radon transform
- density function
- particle filter
- emission computed tomography
- image reconstruction algorithms
- visual features
- maximum a posteriori
- low dimensional
- dimensionality reduction
- principal component analysis
- image representation
- maximum likelihood
- positron emission tomography
- high quality
- high dimensional
- sparsely sampled
- feature vectors
- motion estimation
- riemannian manifolds
- higher order
- markov random field
- natural images