- Integration of CiNii Books functions for fiscal year 2025 has completed
- Trial version of CiNii Research Knowledge Graph Search feature is available on CiNii Labs
- 【Updated on November 26, 2025】Regarding the recording of “Research Data” and “Evidence Data”
- Incorporated Jxiv preprints from JaLC and adding coverage from NDL Search
Restoration of original image from deteriorated image by probabilistic image model
Bibliographic Information
- Published
- 2008-08-01
- DOI
-
- 10.1109/sice.2008.4655196
- Publisher
- IEEE
Description
The conventional noise removal methods are based on spatial filtering and frequency filtering. But these methods have problems associated with degradation of image along side the noise removal. In this study, we propose the method that formulates noise based on multi-dimension Gaussian distribution and restore original image from deteriorated image by Probabilistic inference based on Bayesian statistics. The effectiveness of the proposed method has been validated using benchmark images.
Journal
-
- 2008 SICE Annual Conference
-
2008 SICE Annual Conference 3096-3100, 2008-08-01
IEEE
