Tomographic incoherent holography for microscale X-ray source

Abstract

At present, in the experiment on inertial confinement fusion (ICF), no single imaging diagnosis of the black cavity plasma or the implosion target region can distinguish the emission intensity information in the depth direction, that is, the images acquired by the detector are intensity integral along the detection direction. In this paper, a tomographic imaging method using incoherent holography for microscale X-ray source is introduced. The incoherent holographic imaging technology has an imaging mechanism that encodes and compresses the three-dimensional space information of the light source into a two-dimensional hologram. In the theoretical part, we examine the imaging mechanism of incoherent holographic tomography. Then the compress sensing model which is appropriate for this incoherent tomography is investigated. Combined with the hologram reconstruction algorithm based on compress sensing, the two-dimensional distributions of light intensity at different object distances along the detection direction can be recovered from the two-dimensional hologram. In order to verify the feasibility of this imaging scheme, we simulate the incoherent holographic imaging process of a light source with an axial length of 16 mm, and obtain the tomography light intensity distribution result with a spacing of 4 mm by reconstructing the corresponding incoherent hologram through using the backpropagation algorithms, Wiener filtering algorithm, and compress sensing algorithm. All reconstruction methods mentioned above can recover the corresponding letter light source at a certain object distance, indicating the potential of incoherent holographic technology for three-dimensional imaging. For the backpropagation reconstruction image, there is a large amount of series noise at the edge of the light source signal, which affects signal recognition in practical applications. Although the Wiener filtering algorithm can recognize the image signal to some extent, the low contrast of the reconstructed image results in the distribution of target source strength mixed with background noise. Compared with the algorithm based on the Wiener filtering and backpropagation, compress sensing theory provides a more professional technique for the ill-condition problem. Results from compress sensing reconstruction show that the crosstalk noise is significantly reduced, and the intensity distribution on the objective plane of the light source is basically concentrated in the signal area. The peak-signal-to-noise ratio of reconstructed image is continuously optimized as the number of iterations increases. Besides, the axial and horizontal resolution caused by the innermost ring radius of Fresnel zone plate are also analyzed, indicating that a shorter innermost ring radius can improve the horizontal resolution, bur reduce the axial resolution.