The article describes a new algorithm for the restoration of solar radio images. The technique is based on the use of adaptive regularization procedures that incorporate the k-means clustering algorithm over local roughness measures. Experimental results involving simulated and real images are described. The results demonstrate the superiority of the adaptive procedure compared to conventional regularization, both from the visual and numerical points of view.
Restoration of solar radio images using adaptive regularization techniques based on clustering
2004-01-01
367194 byte
Conference paper
Electronic Resource
English
An Adaptive Segmentation-Based Regularization Term for Image Restoration
British Library Conference Proceedings | 2005
|Adaptive regularization in image restoration by unsupervised learning
British Library Online Contents | 1998
|