Object Specific Trajectory Optimization for Industrial X-ray Computed Tomography

Abstract

In industrial settings, X-ray computed tomography scans are a common tool for inspection of objects. Often the object can not be imaged using standard circular or helical trajectories because of constraints in space or time. Compared to medical applications the variance in size and materials is much larger. Adapting the acquisition trajectory to the object is beneficial and sometimes inevitable. There are currently no sophisticated methods for this adoption. Typically the operator places the object according to his best knowledge. We propose a detectability index based optimization algorithm which determines the scan trajectory on the basis of a CAD-model of the object. The detectability index is computed solely from simulated projections for multiple user defined features. By adapting the features the algorithm is adapted to different imaging tasks. Performance of simulated and measured data was qualitatively and quantitatively assessed.The results illustrate that our algorithm not only allows more accurate detection of features, but also delivers images with high overall quality in comparison to standard trajectory reconstructions. This work enables to reduce the number of projections and in consequence scan time by introducing an optimization algorithm to compose an object specific trajectory.X-ray Computed Tomography (CT) is a widely used tool for inspection in industrial settings, in particular for nondestructive testing (NDT). It is used, for example, for dimensional metrology and defect detection. Apart from the non-existing radiation exposure issue, CT in NDT has several differences when compared to medical CT, such as large variations in object-size and attenuation. One common major issue is the importance of reduced scan times. This facilitates the expansion of industrial CT application from the laboratory to the factory with a full test coverage in the production line.Currently, industrial CT almost exclusively uses standard circular or helical trajectories in combination with filtered backprojection (FBP) reconstruction algorithms. The increased availability of high performance computing hardware, for example GPUs 1,2 , facilitates the revival of iterative reconstruction algorithms, which inherently support arbitrary trajectories. To exploit the flexibility of iterative reconstruction methods, it makes sense to move towards trajectories that include "valuable" acquisition poses (position and orientation of the source/detector arrangement, often also called projections) and exclude "less valuable" acquisition poses. In Varga et al. 3 it was demonstrated that not all acquisition poses have the same value for the reconstructed image. For example, for reconstruction of an edge at least one X-ray has to be tangential to the edge 4 . Furthermore, image quality can be severely reduced by artifacts due to beam-hardening and high attenuation materials. Typically it is not possible to avoid such artifacts completely, but one can choose a trajectory in such a way that the region of interest is not (or less) aff...

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

“…MRI also provides a 3D model of the root (van Dusschoten, Metzner et al 2016) and can be used for time-lapse imaging of growth processes (Jahnke, Menzel et al 2009). The benefits of both X-Ray CT and MRI are substantial and subject to scientific discussion (Fischer, Lasser et al 2016). However, both technologies do not meet the needs for large-scale field studies: Firstly, both technologies restrict plant growth to the size of a given growth container.…”

Registration-based multi-orientation tomography

“…In a CT system where the orientation of the rotation axis can be controlled, an optimal combination of axes orientations can be pre-computed using optimization methods similar to the one described in [30]. In the current work we address the case of a standard laboratory setup, where the rotation axis cannot be adjusted and the user has to simply reposition the specimen on its mount by hand.…”

Optimized multi-shot imaging inspection design

“…The work is complementary to that described in [ 15 ], which relates to the optimal selection of radiographic projections for subsequent tomographic reconstruction, rather than 2D imaging. Key differences between that work and the present study, other than the application, relate to the choice of the means of quantifying the value of a projection and the nature of the applied numerical optimization: the greedy sequential scheme, based on exhaustive enumeration, that Fischer et al .…”