Physically Constrained Iterative Deconvolution
ýBlindţ deconvolution solves for both object f(r) and PSF h(r) simultaneously.
- Ill-posed inverse problem.
- Under ˝ determined: 1 measurement, 2 unknowns
Uses Physical Constraints.
- f(r) & h(r) are positive, real & have finite support.
- Finite support reduces # of variables (symmetry breaking)
- h(r) is band-limited ˝ symmetry breaking
a priori information - further symmetry breaking.
- Noise statistics
- PSF knowledge
- Object & PSF parameterization
- Multiple Frames:
- Same object, different PSFs.
- N measurements, N+1 unknowns.