Processing sequence for non-destructive inspection based on 3D terahertz images

In this paper, a processing sequence has been developed to achieve automated analysis and measurements from 3D terahertz acquisitions. This sequence first performs a tomographic reconstruction from a 3D acquisition made by a continuous millimetre wave system. This reconstruction is based on the Ordered Subsets Convex (OSC) algorithm and has been optimized for 3D THz tomography. Optimisations consist of i) including the Gaussian beam model in the computation of expected measured intensity, ii) modelling the blank scan and dark-field from measurements and, iii) taking account for these models in the algorithm. The new method, called THz-OSC, outperforms both the BFP and THz-SART reconstructions.

Download

Balacey, H., Perraud, J. B., Sleiman, J. B., Guillet, J. P., Recur, B., & Mounaix, P. (2014, November). Processing sequence for non-destructive inspection based on 3D terahertz images. In Infrared, Millimeter-Wave, and Terahertz Technologies III (Vol. 9275, p. 92750J). International Society for Optics and Photonics.

More Stories