Title :
Newton type optimization for Maximum Likelihood blur identification and restoration
Author_Institution :
The Scientific and Technical Research Council of Turkey, Information Technologies Research Institute Gebze, Kocaeli, Türkiye
Abstract :
In this paper, we formulate the blur identification problem as a ML problem and solve by employing a Newton type optimization method based on “The Method of Scoring”. We provide a comparative analysis of EM and Newton type optimization methods.
Keywords :
Convergence; Degradation; Image restoration; Integrated optics; Optical imaging;
Conference_Titel :
Acoustics, Speech, and Signal Processing (ICASSP), 2002 IEEE International Conference on
Conference_Location :
Orlando, FL, USA
Print_ISBN :
0-7803-7402-9
DOI :
10.1109/ICASSP.2002.5745463
