Quantitative Testing of Probabilistic Phase Unwrapping Methods

Computer Science;phase unwrapping;bayesian;probability;testing

Moran, Jodi

University of Waterloo

关键词: Computer Science;    phase unwrapping;    bayesian;    probability;    testing;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/1107/1/jmoran2001.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

The reconstruction of a phase surface from the observed principal values is required for anumber of applications, including synthetic aperture radar (SAR) and magnetic resonanceimaging (MRI). However, the process of reconstruction, called ;;phase unwrapping”, is anill-posed problem. One class of phase-unwrapping algorithms uses smoothness prior modelsto remedy this situation. We categorize this class of algorithms according to the typeof prior model used. Motivated by this categorization, we propose that phase-unwrappingalgorithms be tested by generating phase surfaces from the prior models, and then quantifyingthe deviation of each reconstructed surface from the corresponding original surface.Finally, we present results of the new testing method on a selection of phase-unwrappingalgorithms, including a new algorithm.