"This site requires JavaScript to work correctly"
minimize Header

Prof. Dr.-Ing. Gabriel Herl

Research Robot-CT-Systems

Centre for Applied Research

Technology Campus Plattling CT

Professor

ITC2 0.16

0991/3615-391

Sortierung:
Beitrag in Sammelwerk/Tagungsband

  • Anton Weiss
  • Simon Wittl
  • Gabriel Herl
  • Simon Zabler

Simulated and experimental evaluation of the accuracy of twin robotic CT systems

pg. 1-10.

  • (2023)
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Simon Rettenberger
  • Gabriel Herl
  • Peter Landstorfer
  • Simon Zabler

Multi-Source-CT for inline inspection of extruded profiles

pg. 1-9.

  • (2023)
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Simon Wittl
  • Anton Weiss
  • Gabriel Herl
  • Simon Zabler

Keep Attention to the Mapping: Application of AI for Geometric X-Ray RoboCT Scan Calibration

pg. 1-7.

  • (2023)
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • N. Brierley
  • H. Gaber
  • M. Salzinger
  • K. Chrzan
  • A. Barakat
  • Gabriel Herl

X-ray Computed Tomography for Nuclear Power Plant Maintenance

pg. 1-4.

  • (2023)
  • TC Plattling CT
  • DIGITAL
Zeitschriftenartikel

  • Martin Leipert
  • Gabriel Herl
  • J. Stebani
  • Simon Zabler
  • A. Maier

Three Step Volumetric Segmentation for Automated Shoe Fitting

In: e-Journal of Nondestructive Testing (12th Conference on Industrial Computed Tomography (iCT) 2023, 27 Feb-2 Mar 2023; Fürth, Germany) vol. 28 pg. 1-10.

  • (2023)

DOI: 10.58286/27736

This work presents a three-step segmentation process based on Convolutional Neural Networks. The task is to identify the different parts of shoes from Computed Tomography scans of boxed pairs of shoes. The first step of the three-step algorithm uses a scaled-down volume image to separate the shoe material from its surroundings. The second step segments the shoe's inside volume, i.e. the space enclosed by shoe material. The third and last step splits the segmented shoe material into individual components: shoe upper material, outer and insole. The complete process employs CNNs derived from three-dimensional UNets. Residual SE UNet, Dense UNet, and Bottleneck Residual UNet are evaluated for the three steps. The architectures are modified for large receptive fields. The networks are trained and tested for each step separately and conjointly on CT scans comprising various shoe types. The test results inspire hope for using the process for automated segmentation and extraction of meshes from large batches of CT scans. In particular, the first step using a Residual SE UNet achieves an F1-score of 88.2 % for shoes and 58.9 % for the packing material. The second step segments the inside volume with an F1-score of 81.0 %. The third step segments the shoe into its components and achieves an F1-score for insole of 79.5 %, outer sole of 88.7 % and upper material of 81.3 %.
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Martin Leipert
  • Gabriel Herl
  • Simon Zabler
  • A. Maier

Three Step Volumetric Segmentation for Automated Shoe Fitting

  • (2023)
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Jürgen Wittmann
  • Gabriel Herl

Canny-Net: Known Operator Learning for Edge Detection

pg. 1-7.

  • (2023)
  • TC Plattling CT
  • DIGITAL
Vortrag

  • Gabriel Herl

X-ray CT Data Completeness Condition for Sets of Arbitrary Projections

In: 7th International Conference on Image Formation in X-Ray Computed Tomography (ICIFXCT 2022)

Johns Hopkins University Baltimore, MD, USA

  • 13.06.2022 (2022)
  • TC Plattling CT
  • DIGITAL
Zeitschriftenartikel

  • S. Hatamikia
  • A. Biguri
  • Gabriel Herl
  • G. Kronreif
  • T. Reynolds
  • J. Kettenbach
  • T. Russ
  • A. Tersol
  • A. Maier
  • M. Figl
  • J. Siewerdsen
  • W. Birkfellner

Source-detector trajectory optimization in cone-beam computed tomography: a comprehensive review on today's state-of-the-art

In: Physics in Medicine and Biology vol. 67

  • 16.08.2022 (2022)

DOI: 10.1088/1361-6560/ac8590

Cone-beam computed tomography (CBCT) imaging is becoming increasingly important for a wide range of applications such as image-guided surgery, image-guided radiation therapy as well as diagnostic imaging such as breast and orthopaedic imaging. The potential benefits of non-circular source-detector trajectories was recognized in early work to improve the completeness of CBCT sampling and extend the field of view (FOV). Another important feature of interventional imaging is that prior knowledge of patient anatomy such as a preoperative CBCT or prior CT is commonly available. This provides the opportunity to integrate such prior information into the image acquisition process by customized CBCT source-detector trajectories. Such customized trajectories can be designed in order to optimize task-specific imaging performance, providing intervention or patient-specific imaging settings. The recently developed robotic CBCT C-arms as well as novel multi-source CBCT imaging systems with additional degrees of freedom provide the possibility to largely expand the scanning geometries beyond the conventional circular source-detector trajectory. This recent development has inspired the research community to innovate enhanced image quality by modifying image geometry, as opposed to hardware or algorithms. The recently proposed techniques in this field facilitate image quality improvement, FOV extension, radiation dose reduction, metal artifact reduction as well as 3D imaging under kinematic constraints. Because of the great practical value and the increasing importance of CBCT imaging in image-guided therapy for clinical and preclinical applications as well as in industry, this paper focuses on the review and discussion of the available literature in the CBCT trajectory optimization field. To the best of our knowledge, this paper is the first study that provides an exhaustive literature review regarding customized CBCT algorithms and tries to update the community with the clarification of in-depth information on the current progress and future trends.
  • TC Plattling CT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Gabriel Herl
  • A. Maier
  • Simon Zabler

X-ray CT Data Completeness Condition for Sets of Arbitrary Projections

pg. 23.

  • (2022)

DOI: 10.1117/12.2646435

  • TC Plattling CT
  • DIGITAL
Zeitschriftenartikel

  • Julian Wittmann
  • Gabriel Herl
  • Jochen Hiller

Generation of a 3D model of the inside volume of shoes for e-commerce applications using industrial x-ray computed tomography

In: Engineering Research Express vol. 3 pg. 045058.

  • (2021)

DOI: 10.1088/2631-8695/ac43c8

In 2018, 47% of global internet users had purchased footwear products through the internet, making it the second most popular online shopping category worldwide right after clothing with 57%. In the same year, on average, about every sixth parcel delivered in Germany (16.3%) was returned. With the effort and costs that are associated with the return of shoes, the objective of reducing the number of returns for shoes promises an enormous economic potential and helps to reduce the CO2 emissions due to a lower traffic volume. This paper presents a workflow for determining the inside volume surface of shoes using industrial x-ray computed tomography (CT). The fundamental idea is based on the Region Growing (RG) method for the segmentation of the shoe's inner volume. Experiments are performed to illustrate the correlation of image quality and segmentation result. After obtaining the 3D surface model of an individual foot, the inner volume surface data of a scanned shoe can then be registered and evaluated in order to provide a reliable feedback for the customer regarding the accuracy of fit of a shoe and the individual foot on the basis of an overall 'metric of comfort' before buying online. This step is not part of the work at hand. Conclusions are drawn and suggestions for improving the robustness and the flexibility of the workflow are given, so it can be adapted to various shoe types and implemented in a fully automated measurement process in the future.
  • TC Cham
  • Angewandte Naturwissenschaften und Wirtschaftsingenieurwesen
  • Fraunhofer AWZ CTMT
  • DIGITAL
Zeitschriftenartikel

  • Gabriel Herl
  • Jochen Hiller
  • M. Thies
  • J.-N. Zaech
  • M. Unberath
  • A. Maier

Task-Specific Trajectory Optimisation for Twin-Robotic X-Ray Tomography

In: IEEE Transactions on Computational Imaging vol. 7 pg. 894-907.

  • (2021)

DOI: 10.1109/TCI.2021.3102824

With the advent of robotic C-arm computed tomography (CT) systems in medicine and twin-robotic CT systems in industry, new possibilities for the realisation of complex trajectories for CT scans are emerging. These trajectories will increase the range of CT applications, enable optimisation of image quality for many applications and open up new possibilities to reduce scan time and radiation dose. In this work, trajectory optimisation methods for optimising both, task-based data quality and data completeness, are presented by combining two different metrics. On the one hand, task-based data quality is optimised with a proven observer model. On the other hand, a Tuy-based metric is utilised to optimise data completeness. Both metrics capture mutually exclusive properties of the trajectory which are necessary, but alone are not sufficient for trajectory optimisation. Hence, existing task-driven trajectory optimisation approaches require additional input to decide on an overall optimal trajectory, e.g. in most cases constraints on the trajectory. Advantages and disadvantages of the presented methods are investigated. It is shown that by combining both metrics, trajectory optimisation for arbitrary geometries becomes possible. In application examples it is shown that this can be used for trajectory optimisation of challenging scanning tasks involving metal parts as well as for trajectory optimisation to reduce the number of projections while ensuring task-dependently high image quality. In total, the results of this work enable new applications for X-ray CT, especially for twin-robotic CT systems which are able to benefit from a high number of degrees of freedom.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Zeitschriftenartikel

  • Gabriel Herl
  • Jochen Hiller
  • A. Maier

Scanning trajectory optimisation using a quantitative Tuybased local quality estimation for robot-based X-ray computed tomography

In: Nondestructive Testing and Evaluation vol. (Published 21 June 2020) pg. 1-17.

  • (2020)

DOI: 10.1080/10589759.2020.1774579

Robotic CT systems allow complex scanning trajectories. This work presents a workflow to automatically calculate optimised scanning trajectories for robotic CT systems. In particular, as a local quality estimation, this work introduces a quantitative measure to quantify local reconstruction quality based on the Tuy conditions. The proposed method is tested in two summation experiments using an STL model of a motorcycle. In both experiments, a trajectory is calculated using a quantitative Tuy-based local quality estimation and the reconstruction result is then compared to reconstructions using conventional scanning trajectories. The comparison results indicate that the proposed approach automatically finds trajectories that enable 3D reconstructions with high image quality using much less projection data, which allows a significant reduction of scanning time.
  • Fraunhofer AWZ CTMT
  • Maschinenbau und Mechatronik
  • NACHHALTIG
  • DIGITAL
Zeitschriftenartikel

  • A. Stock
  • Gabriel Herl
  • T. Sauer
  • Jochen Hiller

Edge-preserving compression of CT scans using wavelets

In: Insight - Non-Destructive Testing and Condition Monitoring (The British Institute of Non-Destructive Testing) vol. 62 pg. 345-351.

  • (2020)

DOI: 10.1784/insi.2020.62.6.345

This work addresses the subject of efficient storage of computed tomography (CT) data with an emphasis on the quality of surfaces. Industrial dimensional metrology often requires high measurement accuracy and it is shown that this is retained using wavelet-based compression methods. The applied techniques include a tensor product wavelet transform and soft wavelet shrinkage. In these tests, performed on real objects, dimensional CT measurements of compressed and uncompressed volumes were compared. The necessary storage space was reduced significantly with a negligible loss of accuracy. The storage space required for a multi-sphere phantom was decreased to 4.7% (from 638 MB to 30 MB), with an average deviation below 1 μm from the original volume.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • NACHHALTIG
  • DIGITAL
Zeitschriftenartikel

  • Gabriel Herl
  • Jochen Hiller
  • S. Kasperl
  • A. Maier

Reduktion von Metallartefakten durch multipositionale Datenfusion in der industriellen Röntgen-Computertomographie

In: tm - Technisches Messen vol. 87

  • (2019)

DOI: 10.1515/teme-2019-0137

Metallartefakte stellen eine große Herausforderung für das Messen mit Röntgen-Computertomographie dar. Dieser Beitrag stellt die Methode der multipositionalen Datenfusion zur Reduktion von Metallartefakten vor. Dazu werden mehrere Scans desselben Objekts bei unterschiedlicher Objekt-positionierung durchgeführt, aufeinander registriert und zur Fusion gemeinsam unter Betrachtung eines lokalen Gütemaßes rekonstruiert. In praxisnahen Experimenten wird der Mehrwert der Methode gezeigt. Insbesondere wird dargestellt, wie mit wenig Aufwand und ohne Vorwissen Kunststoffstrukturen trotz starker Metallartefakte sichtbar gemacht werden können, womit das Verfahren ein Alleinstellungsmerkmal gegenüber den existierenden Metallartefaktreduktionsverfahren aufweist.
  • Fraunhofer AWZ CTMT
  • Maschinenbau und Mechatronik
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Gabriel Herl
  • Jochen Hiller
  • T. Sauer

Artifact reduction in X-ray computed tomography by multipositional data fusion using local image quality measures

  • (2019)
Metal artifacts are still a major problem in X-ray industrial computed tomography. In order to reduce metal artifacts and increase the image quality of X-ray CT-scans, we suggest using projection data from multiple scans with differently positioned object orientations. We present two different approaches for multipositional CT, which are especially effective for multimaterial objects with high absorbing metal parts. On one hand, we reconstruct the different scans separatly, estimate the local quality of the resulting volumes and then fuse these volumes to an optimized volume. On the other hand, we introduce smART (shrinking merged Algebraic Reconstruction Technique) and merge sinograms of different scans, estimate the reliability of each projection pixel and then reconstruct the merged sinogram with an adapted SART reconstruction method. We demonstrate our approaches on simulations and on measurement data and are able to show a significant reduction of image artifacts qualitatively and quantitatively with the help of dimensional measurement results.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Peter Landstorfer
  • Gabriel Herl
  • Jochen Hiller

Investigation of Non-circular Scanning Trajectories in Robot-based Industrial X-ray Computed Tomography of Multi-material Objects

  • (2019)
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Vortrag

  • A. Stock
  • Gabriel Herl
  • T. Sauer
  • Jochen Hiller

Edge preserving compression of CT scans using wavelets

In: International Symposium on Structural Health Monitoring and Nondestructive Testing

Saarbrücken

  • 04.-05.10.2018 (2018)
This work addresses the subject of efficient storage of computed tomography (CT) data with an emphasis on the quality of surfaces. Industrial dimensional metrology often requires high measurement accuracy and we show that this is retained using wavelet-based compression methods. The applied techniques include a tensor product wavelet transform and soft wavelet shrinkage. In our tests on real objects, we compared dimensional CT measurements of compressed and uncompressed volumes. We were able to reduce the necessary storage space significantly with a minimal loss of accuracy. For a multi sphere phantom, we decreased the storage space to 4.7% (from 638 MB to 30 MB) with an average deviation below 1 µm from the original volume.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • A. Stock
  • Gabriel Herl
  • T. Sauer
  • Jochen Hiller

Metal artifact reduction by fusion of CT scans from different positions using the unfiltered backprojection

  • (2018)

Metal objects or metal parts in an object are still a major problem of X-ray computed tomography (CT) because of so called metal artifacts. We propose a new method – a multipositional data fusion – for automatically fusing multiple CT volumes from different positions to reduce these metal artifacts. After scanning a specimen several times at different positions and reconstruction of every scan (e.g. by the filtered backprojection), we also perform an unfiltered backprojection. Based on the assumption that metal artifacts occur the most wherever X-rays are attenuated a lot, the unfiltered backprojection is used to autonomously estimate the likelihood of metal artifacts in every voxel of every scan. The different volumes are registered and then fused by weighted sum preferring the voxels with low values in the unfiltered backprojection results. In our tests on real objects, our method fully automatically created optimized volumes with significantly less metal artifacts. The multipositional data fusion was compared to the commercially multi spectra fusion of Werth Messtechnik GmbH and outperformed it in one of the use cases.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • A. Stock
  • Gabriel Herl
  • T. Sauer
  • Jochen Hiller

Edge preserving compression of CT scans using wavelets

  • (2018)

This work addresses the subject of efficient storage of computed tomography (CT) data with an emphasis on the quality of surfaces. Industrial dimensional metrology often requires high measurement accuracy and we show that this is retained using wavelet-based compression methods. The applied techniques include a tensor product wavelet transform and soft wavelet shrinkage. In our tests on real objects, we compared dimensional CT measurements of compressed and uncompressed volumes. We were able to reduce the necessary storage space significantly with a minimal loss of accuracy. For a multi sphere phantom, we decreased the storage space to 4.7% (from 638 MB to 30 MB) with an average deviation below 1 µm from the original volume.
  • Fraunhofer AWZ CTMT
  • Maschinenbau und Mechatronik
  • DIGITAL
Vortrag

  • Gabriel Herl
  • Simon Rettenberger
  • Jochen Hiller
  • T. Sauer

Metal artifact reduction by fusion of CT scans from different positions using the unfiltered backprojection

In: 8th Conference on Industrial Computed Tomography (iCT 2018)

Wels, Österreich

  • 06.-09.02.2018 (2018)
Metal objects or metal parts in an object are still a major problem of X-ray computed tomography (CT) because of so called metal artifacts. We propose a new method – a multipositional data fusion – for automatically fusing multiple CT volumes from different positions to reduce these metal artifacts. After scanning a specimen several times at different positions and reconstruction of every scan (e.g. by the filtered backprojection), we also perform an unfiltered backprojection. Based on the assumption that metal artifacts occur the most wherever X-rays are attenuated a lot, the unfiltered backprojection is used to autonomously estimate the likelihood of metal artifacts in every voxel of every scan. The different volumes are registered and then fused by weighted sum preferring the voxels with low values in the unfiltered backprojection results. In our tests on real objects, our method fully automatically created optimized volumes with significantly less metal artifacts. The multipositional data fusion was compared to the commercially multi spectra fusion of Werth Messtechnik GmbH and outperformed it in one of the use cases.
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL
Beitrag in Sammelwerk/Tagungsband

  • Gabriel Herl
  • Simon Rettenberger
  • Jochen Hiller

Fusion mehrerer Computertomographie-Aufnahmen zur Verbesserung der Bildqualität

pg. 29-32.

  • (2017)
  • Maschinenbau und Mechatronik
  • Fraunhofer AWZ CTMT
  • DIGITAL