A new image reconstruction tool based on non-iterative phase information retrieval from a single diffraction pattern was developed by the group of Prof. Oron.
Lensless imaging techniques enable indirect high resolution observation of objects by measuring the intensity of their diffraction patterns. These techniques utilize radiation in the X-ray regime to image non-periodic objects in sizes that prohibit the use of larger wavelengths. However, retrieving the phase information of the diffraction pattern is not a trivial task, as current methods are divided based on a tradeoff between experimental complexity and computational reconstruction efficiency.
The method described here is suitable for use with existing lensless imaging techniques to provide direct, robust and efficient phase data while requiring reduced computational and experimental complexity. This method, demonstrated in a laboratory setup on 2D objects, is also applicable in 1D. It can be applied to various phase retrieval applications such as coherent diffractive imaging and ultrashort pulse reconstruction
- Phase microscopy
- Signal processing
- X-ray imaging
- A Generic solution to the phase retrieval problem
- Non-iterative approach
- An efficient and noise robust tool
The method is based on the fact that the Fourier transform of the diffraction intensity measurement is the autocorrelation of the object. The autocorrelation and cross-correlations of two sufficiently separated objects are spatially distinct. Based on this, the method consists of three main steps: (a) The sum of the objects autocorrelations, as well as their cross-correlation, are reconstructed from the Fourier transform of the measured diffraction pattern. (b) The individual objects autocorrelations are reconstructed from their sum and the cross-correlation. (c) Using the two intensities and the interference cross term, double-blind Fourier holograph is applied to recover the phase by solving a set of linear equations.