Related Experiment Videos
Retrieval of a complex-valued object from its diffraction pattern
W McBride1, N L O'Leary, L J Allen
1School of Physics, University of Melbourne, Victoria 3010, Australia.
Physical Review Letters
|December 17, 2004
Summary
Object retrieval from diffraction patterns is challenging, especially with symmetric support constraints. This study demonstrates an iterative algorithm using a difference map effectively reconstructs complex-valued objects with symmetric supports, requiring oversampling and careful parameter selection.
Area of Science:
- Diffraction imaging
- Computational imaging
- Image reconstruction
Background:
- Object retrieval from diffraction patterns is a complex inverse problem.
- Symmetric support constraints often exacerbate reconstruction difficulties.
- Iterative algorithms are commonly employed but may face convergence issues.
Purpose of the Study:
- To present an effective iterative algorithm for complex-valued object retrieval from diffraction patterns.
- To address the specific challenge of symmetric support constraints in diffraction imaging.
- To demonstrate routine convergence for such difficult cases.
Main Methods:
- Utilizing an iterative algorithm based on a difference map.
- Applying the algorithm to cases with symmetric support constraints.
- Employing sufficient oversampling in the diffraction pattern data.
Main Results:
- The proposed iterative algorithm demonstrates routine convergence for complex-valued object retrieval.
- Successful reconstruction is achieved even with symmetric support constraints.
- Judicious selection of difference map parameters is crucial for convergence.
Conclusions:
- An iterative difference map algorithm provides a robust solution for object retrieval with symmetric supports.
- Oversampling and parameter tuning are key factors for successful diffraction-based reconstruction.
- This method offers a reliable approach to a previously difficult imaging problem.