Single-Image Geometric-Flow X-Ray Speckle Tracking

Abstract

We develop a speckle-tracking method for x-ray phase-contrast imaging, based on the concept of geometric flow. This flow is a conserved current associated with deformation of illuminating x-ray speckles induced by passage through a sample. The method provides a rapid, efficient, and accurate algorithm for quantitative phase imaging. It is highly photon efficient and able to image dynamic objects, since a single radiograph of the sample is sufficient for the phase recovery. We experimentally quantify the resolution and contrast of the approach with both two-dimensional and three-dimensional phase-imaging applications using x-ray synchrotron radiation. Finally, we discuss adaptations of the method to imaging with compact x-ray sources that have a large source size and significant spectral bandwidth.

arXiv comment3 figures

References in corpus (2)

• PyHST2: an hybrid distributed code for high speed tomographic reconstruction with iterative reconstruction and a priori knowledge capabilities
• Near-field speckle-scanning-based x-ray imaging

Cited by in corpus (8)

• X-ray Fokker--Planck equation for paraxial imaging
• Single-shot x-ray speckle-based imaging of a single-material object
• X-ray Multimodal Intrinsic-Speckle-Tracking
• Paraxial diffusion-field retrieval
• X-ray dark-field via spectral propagation-based imaging
• Directional dark-field implicit x-ray speckle tracking using an anisotropic-diffusion Fokker-Planck equation
• Speckle-based X-ray microtomography via preconditioned Wirtinger flow
• Low-exposure, high-quality multimodal speckle X-ray imaging via an intrinsic gradient-flow approach