Depth map super-resolution via Markov Random Fields without texture-copying artifacts.
Kai-Han LoKai-Lung HuaYu-Chiang Frank WangPublished in: ICASSP (2013)
Keyphrases
- depth map
- super resolution
- markov random field
- high quality
- high resolution
- low resolution
- image restoration
- mrf models
- depth information
- super resolved
- image reconstruction
- d scene
- belief propagation
- random fields
- mrf model
- energy function
- higher order
- graph cuts
- potential functions
- stereo matching
- pairwise
- depth estimation
- motion estimation
- maximum a posteriori
- low resolution images
- single image
- energy minimization
- image segmentation
- multi frame
- image super resolution
- light field
- motion blur
- image processing
- spatial resolution
- feature extraction
- high resolution images
- three dimensional
- edge preserving