Convex Generalizations of Total Variation Based on the Structure Tensor with Applications to Inverse Problems.
Stamatios LefkimmiatisAnastasios RoussosMichael UnserPetros MaragosPublished in: SSVM (2013)
Keyphrases
- inverse problems
- structure tensor
- anisotropic diffusion
- convex optimization
- partial differential equations
- total variation
- nonlinear diffusion
- global optimization
- denoising
- diffusion tensor
- edge preserving
- optical flow
- image reconstruction
- image denoising
- corner detection
- noise removal
- image processing
- level set
- edge detection
- image restoration
- multiscale
- early vision
- feature selection
- image enhancement
- optimization methods
- positive definite
- optimization problems
- human vision
- smoothness constraint
- parameter space
- graphical models
- kl divergence
- noisy images
- image quality