Title :
New representation of TV regularization for super-resolution
Author :
Fazli, Saeid ; Bouzari, Hamed ; Pour, Hamed Moradi
Author_Institution :
Electr. Eng. Dept., Zanjan Univ., Zanjan, Iran
Abstract :
Total Variation (TV) Regularization acts like an anisotropic diffusion while in smooth areas it has isotropic behavior, which makes it very useful for recovering edges in images and at the same time having no effect on smooth areas but it suffers from high non-linearity. This paper presents a new method to handle this problem. The proposed method for Super-Resolution and image Inpainting is based on TV regularization. Experimental results show the effectiveness and high performance of the method.
Keywords :
computer vision; image resolution; anisotropic diffusion; image inpainting; isotropic behavior; super-resolution; total variation regularization; Gold; Image resolution; Image restoration; Manganese; Pixel; Signal resolution; TV; image inpainting; super-resolution; total variation;
Conference_Titel :
GCC Conference & Exhibition, 2009 5th IEEE
Conference_Location :
Kuwait City
Print_ISBN :
978-1-4244-3885-3
DOI :
10.1109/IEEEGCC.2009.5734291
