Deriving instrumental point spread functions from partially occulted images.

Area of Science:

• Optical imaging
• Image processing
• Instrument calibration

Background:

• The point spread function (PSF) is crucial for understanding imaging system performance.
• Accurate PSF determination is key to correcting image aberrations like scattering and defocus.
• Existing methods may require specific image types or predefined PSF models.

Purpose of the Study:

• To develop a novel semi-empirical, semi-blind method for deriving the PSF from partially occulted images.
• To enable accurate PSF reconstruction without assuming a specific functional form.
• To improve image spatial resolution and dynamic contrast.

Main Methods:

• Partitioning the 2D PSF into multiple segments.
• Formulating a multilinear system of equations.
• Directly fitting the system to determine segment weights for PSF reconstruction.

Main Results:

• The developed algorithm converges to the true instrumental PSF for various occultation types.
• The method does not require a predefined PSF functional form.
• Achieved high accuracy: central PSF weight within 1.2%, overall PSF error typically 0.5-5%.

Conclusions:

• The presented methodology offers a robust and versatile approach to PSF derivation.
• Applicable to diverse imaging scenarios, including laboratory, industrial, and astronomical settings.
• Significantly improves the fidelity of reconstructed images by accurately characterizing optical system behavior.