Skip to Main content Skip to Navigation
Conference papers

Bayesian framework for solving transform invariant low-rank textures

Abstract : As a holistic image feature, Transform Invariant Low-Rank Textures (TILT) can effectively recover the rectification of user-specified patch with a rich class of low-rank textures in 2D images. Unlike conventional local image features, TILT isn't dependent on the extraction of points, corners or edges, which would bring inaccuration and weak robustness. However , TILT is still rather rudimentary, and have some limitations in applications. In this paper, we proposed a novel algorithm for better solving TILT. Our method is based on the application of Bayesian framework in robust principal component analysis (RPCA), some missing entries of TILT's mathematical model are taken into account as well. Experimental results on both synthetic and real data indicate that our new algorithm outperforms the existing algorithm especially for the case with corruptions and occlusions.
Document type :
Conference papers
Complete list of metadata

Cited literature [12 references]  Display  Hide  Download
Contributor : Lei Yu Connect in order to contact the contributor
Submitted on : Friday, December 2, 2016 - 4:27:35 AM
Last modification on : Tuesday, February 27, 2018 - 2:03:58 PM


Files produced by the author(s)



Shihui Hu, Lei Yu, Menglei Zhang, Chengcheng Lv. Bayesian framework for solving transform invariant low-rank textures. ICIP, 2016, Phinex, United States. pp.3588 - 3592, ⟨10.1109/ICIP.2016.7533028⟩. ⟨hal-01407365⟩



Record views


Files downloads