João Paulo Papa,  Greice Martins de Freitas. 

Iterative image restoration algorithms can better control the restoration process, but the choice of the relaxation parameters can be an inviable task. We present here a GP-based approach that stabilizes the iterative RAP-2D image restoration algorithm choosing the optimal or quasioptimal relaxation parameters that maximize the improvement signal to noise ratio of phantom images. CBERS-2 CCD band 2 satellite images were also restored to validate the proposed GP-based framework for image restoration.

http://www.lbd.dcc.ufmg.br/colecoes/wvc/2008/0025.pdf

Processando consulta.