Common types of artifacts in industrial computed tomography (CT) and their improvement methods.
Industrial CT plays a significant role in non-destructive testing, scientific research, and other fields, providing tomographic images of the inspected objects and revealing information such as the character, shape, position, and size of the defects. However, due to inherent factors such as radiation scattering, unstable dose, and data acquisition errors, artifacts often appear in the original CT images, directly affecting the measurement accuracy. By using specialized image acquisition and reconstruction software, calculating and correcting image errors, effectively calibrating inconsistent response areas of the detectors and the complex structure of the inspected objects to reduce imaging deviations, thereby improving imaging quality.
Common artifact types and improvement methods
1. Ring artifacts
Definition: Ring artifacts are a common type of artifact in industrial CT scanning, caused by detector pixel channel damage, non-linearity and inconsistency of radiation intensity and radiation energy spectrum response. In the reconstructed image, they appear as a series of concentric circles or arcs centered on the rotation center.
Improvement methods:Specifically adjust the performance of the detector, improve 3D X-ray CT inspection system design or optimize the image reconstruction algorithms to effectively reduce the impact of ring artifacts.
2. Metal Artefacts
Definition: Metal artefacts mainly result from the physical properties when 3D X-ray CT inspection system scanning the metallic materials object, such as absorption, scattering, and reflection of radiation. Metal artefacts can cause problems such as uneven brightness, blurred edges, or radial stripes in the reconstructed images.
Improvement methods:Various methods such as interpolation correction, iterative correction, and hybrid correction adjust the projection data or the reconstruction algorithm to improve the artefact issue.
3. Beam hardening artifacts
Definition: Beam hardening artifacts are caused by the fact that low-energy photons are absorbed faster than high-energy photons when the X-ray beam passes through the object, resulting in an increase in the "hardness" of the X-ray beam. This artifact can cause some striped or dark band-like artifacts without obvious discontinuities or clear boundaries in the reconstructed image.
Improvement methods:Using traditional methods such as software correction and iterative reconstruction, algorithms based on deep learning can automatically detect and remove artifacts by learning from a large amount of normal sample data.
Summary
These improvement methods can significantly reduce the impact of artifacts on the test results, improve the imaging quality and the reliability of the detection results. That can provide strong support for 3D X-ray CT inspection system design and apply on three-dimensional non-destructive analysis projects, such as wall thickness measurement, pore and crack analysis, size measurement, CAD comparison, and reverse engineering.