Title :
A fast algorithm for maximum-likelihood imaging with coherent speckle measurements
Author :
Schulz, Timothy J.
Author_Institution :
Dept. of Electr. Eng., Michigan Technol. Univ., Houghton, MI, USA
Abstract :
The application of the space alternating, expectation-maximization (SAGE) method is discussed for a problem in coherent-speckle imaging (or two-dimensional spectrum estimation). For this algorithm, convergence of the likelihood is demonstrated to be significantly faster than that of an expectation-maximization (EM) based algorithm that has been previously proposed. In addition, the utilization of the SAGE method for penalized maximum-likelihood estimation is demonstrated
Keywords :
Gaussian processes; array signal processing; convergence of numerical methods; direction-of-arrival estimation; image processing; maximum likelihood estimation; random processes; speckle; spectral analysis; 2D spectrum estimation; Gaussian random process; SAGE method; algorithm convergence; coherent speckle measurements; coherent-speckle imaging; expectation-maximization based algorithm; maximum-likelihood imaging; penalized maximum-likelihood estimation; sensors array; space alternating expectation-maximization; Image sensors; Laser radar; Maximum likelihood estimation; Radar imaging; Radar scattering; Sensor arrays; Sensor phenomena and characterization; Space technology; Speckle; Spectral analysis;
Conference_Titel :
Image Processing, 1997. Proceedings., International Conference on
Conference_Location :
Santa Barbara, CA
Print_ISBN :
0-8186-8183-7
DOI :
10.1109/ICIP.1997.648004
