Structure tensor adaptive total variation for image restoration

Image denoising and restoration is one of the basic requirements in many digital image processing systems. Variational regularization methods are widely used for removing noise without destroying edges that are important visual cues. This paper provides an adaptive version of the total variation regularization model that incorporates structure tensor eigenvalues for better edge preservation without creating blocky artifacts associated with gradient-based approaches. Experimental results on a variety of noisy images indicate that the proposed structure tensor adaptive total variation obtains promising results and compared with other methods, gets better structure preservation and robust noise removal.