A content-aware image prior
n image restoration tasks, a heavy-tailed gradient distribution of natural images has been extensively exploited as an image prior. Most image restoration algorithms impose a sparse gradient prior on the whole image, reconstructing an image with piecewise smooth characteristics. While the sparse gra...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2012
|
| Online Access: | http://hdl.handle.net/1721.1/71890 https://orcid.org/0000-0002-2231-7995 |
| _version_ | 1811069204226375680 |
|---|---|
| author | Cho, Taeg Sang Joshi, Neel Zitnick, C. Lawrence Kang, Sing Bing Szeliski, Richard Freeman, William T. |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Cho, Taeg Sang Joshi, Neel Zitnick, C. Lawrence Kang, Sing Bing Szeliski, Richard Freeman, William T. |
| author_sort | Cho, Taeg Sang |
| collection | MIT |
| description | n image restoration tasks, a heavy-tailed gradient distribution of natural images has been extensively exploited as an image prior. Most image restoration algorithms impose a sparse gradient prior on the whole image, reconstructing an image with piecewise smooth characteristics. While the sparse gradient prior removes ringing and noise artifacts, it also tends to remove mid-frequency textures, degrading the visual quality. We can attribute such degradations to imposing an incorrect image prior. The gradient profile in fractal-like textures, such as trees, is close to a Gaussian distribution, and small gradients from such regions are severely penalized by the sparse gradient prior. To address this issue, we introduce an image restoration algorithm that adapts the image prior to the underlying texture. We adapt the prior to both low-level local structures as well as mid-level textural characteristics. Improvements in visual quality is demonstrated on deconvolution and denoising tasks. |
| first_indexed | 2024-09-23T08:07:22Z |
| format | Article |
| id | mit-1721.1/71890 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T08:07:22Z |
| publishDate | 2012 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/718902022-09-30T07:39:35Z A content-aware image prior Cho, Taeg Sang Joshi, Neel Zitnick, C. Lawrence Kang, Sing Bing Szeliski, Richard Freeman, William T. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Freeman, William T. Cho, Taeg Sang Freeman, William T. n image restoration tasks, a heavy-tailed gradient distribution of natural images has been extensively exploited as an image prior. Most image restoration algorithms impose a sparse gradient prior on the whole image, reconstructing an image with piecewise smooth characteristics. While the sparse gradient prior removes ringing and noise artifacts, it also tends to remove mid-frequency textures, degrading the visual quality. We can attribute such degradations to imposing an incorrect image prior. The gradient profile in fractal-like textures, such as trees, is close to a Gaussian distribution, and small gradients from such regions are severely penalized by the sparse gradient prior. To address this issue, we introduce an image restoration algorithm that adapts the image prior to the underlying texture. We adapt the prior to both low-level local structures as well as mid-level textural characteristics. Improvements in visual quality is demonstrated on deconvolution and denoising tasks. United States. National Geospatial-Intelligence Agency (NEGI- 1582-04-0004) United States. Army Research Office. Multidisciplinary University Research Initiative. (Grant Number N00014-06-1-0734) Samsung Scholarship Foundation MIT Summer Research Program (Internship) 2012-07-30T16:30:40Z 2012-07-30T16:30:40Z 2010-08 2010-06 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-6984-0 1063-6919 http://hdl.handle.net/1721.1/71890 Cho, Taeg Sang et al. “A Content-aware Image Prior.” IEEE, 2010. 169–176. © Copyright 2010 IEEE https://orcid.org/0000-0002-2231-7995 en_US http://dx.doi.org/ 10.1109/CVPR.2010.5540214 2010 IEEE Conference on Computer Vision and Pattern Recognition Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers (IEEE) IEEE |
| spellingShingle | Cho, Taeg Sang Joshi, Neel Zitnick, C. Lawrence Kang, Sing Bing Szeliski, Richard Freeman, William T. A content-aware image prior |
| title | A content-aware image prior |
| title_full | A content-aware image prior |
| title_fullStr | A content-aware image prior |
| title_full_unstemmed | A content-aware image prior |
| title_short | A content-aware image prior |
| title_sort | content aware image prior |
| url | http://hdl.handle.net/1721.1/71890 https://orcid.org/0000-0002-2231-7995 |
| work_keys_str_mv | AT chotaegsang acontentawareimageprior AT joshineel acontentawareimageprior AT zitnickclawrence acontentawareimageprior AT kangsingbing acontentawareimageprior AT szeliskirichard acontentawareimageprior AT freemanwilliamt acontentawareimageprior AT chotaegsang contentawareimageprior AT joshineel contentawareimageprior AT zitnickclawrence contentawareimageprior AT kangsingbing contentawareimageprior AT szeliskirichard contentawareimageprior AT freemanwilliamt contentawareimageprior |