In order to encourage industrial interest in terahertz systems for non-destructive testing, this thesis features various methods and techniques for improving the imaging and thickness measurement performance of sub-THz FMCW (Frequency Modulated Continuous Wave) radar sensors.
Focused imaging approaches are restricted by diffraction and involve a compromise between lateral resolution and depth of field.
The SAR (Synthetic Aperture Radar) imaging technique overcomes these limitations. In this context, a new algorithmic approach is proposed, significantly improving lateral resolution by a factor of three compared with conventional focused imaging systems.
Point-by-point acquisition systems are often limited in measurement speed due to the constraints associated with translation vectors, which are subject to excessively high inertias.
To tackle this problem, an ultra-fast imaging system has been developed. This incorporates a rotating polygonal mirror which deflects the radar beam and directs it onto a telecentric f-theta lens, which focuses the beam onto a plane.
In addition, sub-THz FMCW radar transceivers are not limited to non-destructive imaging, but also enable precise measurement of material thickness. However, these measurements are usually limited by the longitudinal resolution of the radar unit.
This study proposes an innovative algorithm, combined with a new genetic algorithm approach, which overcomes this limitation by measuring thicknesses five times thinner than the longitudinal resolution in the air of the sensor.
