Total variation norm-based nonnegative matrix factorization for identifying discriminant representation of image patterns.
Taiping ZhangBin FangWeining LiuYuan Yan TangGuanghui HeJing WenPublished in: Neurocomputing (2008)
Keyphrases
- total variation
- blind deconvolution
- fidelity term
- nonnegative matrix factorization
- total variation regularization
- image representation
- denoising
- image restoration
- image denoising
- regularization term
- image features
- image data
- image classification
- input image
- convex optimization
- minimization problems
- multiscale
- matrix factorization
- interior point
- image deblurring
- regularization methods
- objective function
- negative matrix factorization
- image processing
- least squares
- markov random field
- feature extraction
- restored image
- augmented lagrangian
- computer vision
- natural images
- image segmentation