Tomographic Image Deblurring via Position-Dependent PSF Maps

The patent applies preliminary action by acquiring a point spread function (PSF) map before image reconstruction. The PSF map characterizes the blurring effects that will occur during scanning, allowing the system to pre-calculate deblurring parameters. This preliminary characterization enables subsequent deblurring processing to effectively compensate for motion artifacts, resolving the contradiction between obtaining clear images and avoiding blurring artifacts during scanning.

The patent applies local quality by using a position-dependent PSF map where different regions of the field of view have customized deblurring parameters. Instead of applying uniform deblurring across the entire image, the system tailors the deblurring strength and characteristics to local regions, allowing aggressive deblurring where needed while preserving noise-free regions. This localized approach resolves the contradiction by enabling selective deblurring that improves edge clarity without uniformly boosting noise across the image.

The patent applies segmentation by dividing the field of view into multiple regions, each with its own PSF characteristics stored in a PSF map. Rather than using a single global PSF, the system segments the imaging space and assigns position-specific PSF parameters to each region. This segmentation enables accurate deblurring that accounts for spatial variations in blurring effects while maintaining manageable processing complexity through efficient map-based parameter storage and retrieval.