Terahertz diffractive imaging with saturated data inpainting

This article focuses on improving diffraction imaging techniques in the field of Térahertz frequencies, by approaching the problem of data saturation due to the limited dynamic range of detection systems.

We introduce with the ITMO team (Nikolay V Petrov Team) a new approach to algorithmic reconstruction which deals with saturated areas in the measured intensity data sets, allowing reliable recovery of the complex amplitude of the waves. This technique, applicable in various spectral ranges, has been tested specifically in the Térahertz frequency range, where the problems linked to sources and detectors are particularly acute. They demonstrate that the amplitude and phase distributions recovered are comparable in quality to those reconstructed from reference techniques with a dynamic high range.

The document explores the use of Terahertz phase imaging techniques with continuous waves sources, widely adopted in industrial applications because of their experimental facility and their non -complex detection system. We discuss the importance of information on the phase delay provided by these techniques to determine the topographic properties of samples, the hidden structures of optically opaque dielectric objects and for tomographic reconstruction.

The article also examines the different CW phase measurement techniques applied with THZ radiation, such as digital holography, interferometry, intensity transport equation and phase iterative reprocessing (IPR), in Emphasizing the approaches of pttychography and multiple algorithms like the single Beam multiple intensity reconstruction (SBMIR).

Reconstructed phase (top row) and amplitude (middle row) distributions by SBMIR (a, d, c, f) and SBMIR-I (b, e) from saturated (a, b, d, e); and HDR (c, f) datasets. The bot tom row displays NRMSEobj (j), showing the performance of SBMIR and SBMIR-I algorithms on the saturated dataset (g); histograms (h) of the amplitudes, displayed in (e) and (f), and their difference |uobj − uobjHDR | (i)

One of the main challenges covered is the low dynamic interval of the detection systems used in the Thz field, which causes saturation problems. The proposed solution aims to simplify the data acquisition process and expand possibilities in the design of measurement tools and the study of dynamic scenes.

Link to the pdf document

Tsiplakova, E. G., Perraud, J. B., Chopard, A., Guillet, J. P., Mounaix, P., & Petrov, N. V. (2023). Terahertz diffractive imaging with saturated data inpainting. Optics Letters48(21), 5463-5466.


More Stories