Spherically symmetric volume elements as basis functions for image reconstructions in computed laminography.

Area of Science:

• Tomographic imaging
• Image reconstruction algorithms
• Scientific computing

Background:

• Laminography is a tomographic technique for 3D imaging of large, flat objects.
• Iterative algorithms, like the Kaczmarz method, are used for laminography reconstruction.

Purpose of the Study:

• Develop a superior reconstruction algorithm for challenging laminography problems.
• Enhance image quality in computed laminography.

Main Methods:

• Replaced traditional voxel basis functions with spherically symmetric volume elements (blobs).
• Utilized generalized Kaiser-Bessel window functions for blob basis functions.
• Adapted the algebraic reconstruction technique (ART) for blob-based reconstruction.

Main Results:

• Blob basis functions improved resolution (FWHM) from 10.2±1.0 to 9.9±0.9 (p=2.3·10-4).
• Signal-to-noise ratio increased from 16.1 to 31.0.
• Faster convergence compensated for increased computational cost per iteration.

Conclusions:

• Blob basis functions offer superior tomographic reconstruction for laminography.
• Recommended for datasets with expected noise or limited projections.
• Improves image quality and data analysis in computed laminography.