Multi-Material Empirical Artifact Correction (MEAK): Efficient Algorithm for Correcting Beam Hardening Artifacts in Multi-Material CT

Industrial X-ray computed tomography (CT) has become a well-established method for the fast, complete, and non-destructive measurement of workpieces. However, in multi-material components containing highly absorbing materials, beam hardening artifacts result in systematic measurement errors [1]. Existing approaches to artifact reduction, such as the use of hard prefilters and multi-spectral CT (MSP-CT) methods [2], require additional measurement time and reduce the contrast-to-noise ratio (CNR) [3]. This contribution introduces the Multi-Material Empirical Artifact Correction (MEAK), a novel correction method that reduces beam hardening artifacts without increasing measurement time and can be seamlessly integrated into existing processes. Comparative measurements demonstrate that MEAK reduces both linear and non-linear artifacts, preserves the CNR, and halves the measurement time compared to MSP-CT.