Super-resolution/segmentation of 2D trabecular bone images by a Mumford-Shah approach and comparison to total variation.
Y. LiAlina TomaBruno SixouFrançoise PeyrinPublished in: EUSIPCO (2016)
Keyphrases
- mumford shah
- total variation
- super resolution
- image restoration
- piecewise smooth
- variational framework
- blind deconvolution
- low resolution
- mumford shah functional
- degraded images
- gaussian noise
- image deblurring
- multi frame
- light field
- image segmentation
- high resolution
- denoising
- piecewise constant
- image reconstruction
- bayesian estimation
- segmentation method
- segmentation algorithm
- motion estimation
- regularization term
- image processing
- edge preserving
- level set
- high quality
- image denoising
- test images
- level set method
- object recognition
- motion blur
- edge detection
- energy functional
- regularization parameter
- curve evolution
- shape prior
- variational methods
- reconstructed image
- sparse representation
- input image
- blurred images