Optimal Measurement Configuration in Computational Diffractive Imaging

TL;DR

This paper introduces an optimized measurement configuration method for computational diffractive imaging that enhances spectral reconstruction quality by using a sequential backward selection algorithm to identify configurations minimizing expected error.

Contribution

It adapts a sequential backward selection algorithm for optimal measurement configuration in diffractive imaging, reducing complexity through system approximation and assumptions.

Findings

Optimized configurations significantly improve reconstruction performance.

The method reduces computational complexity compared to exhaustive search.

Numerical results demonstrate superior accuracy over standard configurations.

Abstract

Diffractive lenses have recently been applied to the domain of multispectral imaging in the X-ray and UV regimes where they can achieve very high resolution as compared to reflective and refractive optics. Conventionally, spectral components are reconstructed by taking measurements at the focal planes. However, the reconstruction quality can be improved by optimizing the measurement configuration. In this work, we adapt a sequential backward selection algorithm to search for a configuration which minimizes expected reconstruction error. By approximating the forward system as a circular convolution and making assumptions on the source and noise, we greatly reduce the complexity of the algorithm. Numerical results show that the configuration found by the algorithm significantly improves the reconstruction performance compared to a standard configuration.