|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.stsci.edu/stsci/meetings/irw/proceedings/nunezj.dir/section3_6.html
Дата изменения: Mon Apr 18 23:42:54 1994 Дата индексирования: Sun Dec 23 19:34:22 2007 Кодировка: Поисковые слова: п п п п п п п п п п п п п п |
In this paper we have developed a variant of the Bayesian algorithm with entropy prior (FMAPE) that uses a balancing parameter that is variant in space. To the desirable characteristics of the FMAPE algorithm, this development adds the capability of different degrees of smoothing in different regions of the image. This implies a variation in resolution in the restoration, allowing high resolution in high S/N regions and lower resolution with lower artifacts in poor S/N regions
The algorithm can prevent amplification of noise while maintaining the photometric accuracy in the whole image (background, diffuse objects and stars).
We have applied the algorithm to the planetary nebula simulation using three different values for the hyperparameter corresponding to the background, nebula, and stars respectively, with excellent results both in the general aspect of the restoration and in the photometry.
More work is still needed to determine the different hyperparameters to use in the different regions. We are studying image segmentation techniques to optimize the process.