Dynamic angle selection for few-view X-ray inspection of CAD based objects

Presenti, Alice; Sijbers, Jan; Beenhouwer, Jan De

doi:10.1117/12.2534894

Cited by 3 publications

(2 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the 3D object pose is often only approximately known. In our previous work [5], 3D spatial information of the object (orientation and position) was estimated by mathematically joining 2D information from multiple projections, and the system was driven to acquire task specific projections. However, the method is limited to cylindrical objects.…”

Section: Introductionmentioning

confidence: 99%

Deep learning-based 2D-3D sample pose estimation for X-ray 3DCT

Presenti¹,

Bazrafkan²,

Sijbers³

et al. 2020

eJNDT

View full text Add to dashboard Cite

3D X-ray Computed Tomography (CT) is increasingly being used for non-destructive inspection of objects. Conventional CT inspection requires many projections, typically spanning 360 to reconstruct a 3D image of the object, which is then segmented and subsequently compared with the reference computer-aided design (CAD) model. Such an inspection flowchart, however, is a time inefficient procedure, not suitable for inline inspection. To overcome this problem, we directly compare the measured projections with simulated ones from the CAD model. To do so, the simulated projections need to be created with the same acquisition geometry as the measured ones. When an object is inserted on a scanning system, its orientation may vary with respect to the default CAD model orientation. For this reason, 2D/3D registration between the CAD model and the measured projections of the real object is necessary. In this paper, we present a deep learning based method to accurately estimate the 3D orientation of an object from one projection image.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep learning-based 2D-3D sample pose estimation for X-ray 3DCT

Presenti¹,

Bazrafkan²,

Sijbers³

et al. 2020

eJNDT

View full text Add to dashboard Cite

show abstract

“…In [38], Haque et al proposed projection selection methods based on reconstruction error and spectral richness, and have shown an example where they used entropy-based uncertainty. The authors of [91] used the structural similarity index measure [128] to estimate the sample rotation using a CAD projector [72] toolbox. Dabravolski et al [21] proposed a dynamic algorithm, in which new projection angles are selected by maximizing the information gain about the object, given the set of possible new angles.…”

Section: Introductionmentioning

confidence: 99%

Image Quality Improvement in Computed and Binary Tomography

Lékó

View full text Add to dashboard Cite

First and foremost, I would like to express my deep gratitude to my supervisor, Péter Balázs, who guided me throughout my PhD years. I will always be thankful to him and will never forget the great and joyful moments we have spent together during our journey. I wish to extend my special thanks to László Varga for his invaluable assistance that he provided day-by-day and for offering his friendship from the first day we met. Without them, this thesis could never have been completed.I would like to thank the following people for helping me finalise my PhD studies:

show abstract

CNN-based Pose Estimation of Manufactured Objects During Inline X-ray Inspection

Presenti

Liang

Pereira

et al. 2021

2021 IEEE 6th International Forum on Research and Technology for Society and Industry (RTSI)

Self Cite

View full text Add to dashboard Cite

X-ray Computed Tomography (CT) is widely used in inspection of manufactured objects, but typically requires hundreds of radiographs over a wide angular range, making it unsuitable for real-time applications and limited angular view imaging. Fortunately, inspection can also be performed by directly comparing measured radiographs with simulated ones from a reference model. For an effective comparison, an accurate alignment between the radiographs is crucial. In this work, we propose a deep learning-based 3D pose estimation of the object to be inspected from a single radiograph.

show abstract

Dynamic angle selection for few-view X-ray inspection of CAD based objects

Cited by 3 publications

References 3 publications

Deep learning-based 2D-3D sample pose estimation for X-ray 3DCT

Deep learning-based 2D-3D sample pose estimation for X-ray 3DCT

Image Quality Improvement in Computed and Binary Tomography

CNN-based Pose Estimation of Manufactured Objects During Inline X-ray Inspection

Contact Info

Product

Resources

About