VOM2MPS: From Virtual Outcrop Models to Multi-Point Statistics Training Images for Improved Reservoir Modelling

Reservoir modelling of sedimentary systems is important to understand subsurface geology, and is widely used in the oil and gas industry, CO2 capture and storage and aquifers. The models can be constructed based on facies interpretations using outcrop analogues, which are made available in databases such as SAFARI. Geostatistical modelling has been recently complemented by multipoint statistics (MPS). MPS offer advantages over traditional techniques, but its application is hampered by the need for case-specific, volumetric training images (TI). TI creation demands a rapid interpretation of virtual outcrop models (VOMs) and new workflows for doing so.

VOM2MPS is a knowledge-building research project that addresses gaps in the rapid interpretation of outcrop models, the creation of MPS Tis and industry-applicable modelling workflows for generating and working with TIs. It builds on a large number of collected virtual outcrops within the SAFARI project. VOM2MPS is split into two interlinked parts, actively being addressed by one PhD candidate each. The project leading, coordination and student supervision is provided by Dr. Simon J. Buckley (Uni Research) and Prof. John A. Howell (University of Aberdeen).

James R. Mullins (Aberdeen) works on geological research aspects, aiming at establishing an industry-applicable workflow for MPS TI generation from outcrop interpretations and virtual outcrops. The TIs are subsequently used in stochastic MPS modelling procedures. A library of training images, linked to SAFARI, is prepared from the outcrop studies conducted during the research period. The scientific research focuses on large-scale, stochastic reservoir modelling to study reservoir uncertainties caused by heterogeneities. The project is investigating other TI sources, such as process-based models and sediment simulation outputs.

Christian Kehl (Bergen) is working on computer-scientific and geomatics research aspects, researching novel- and refining existing algorithms for the rapid interpretation and facies mapping of virtual outcrop models. The techniques are implemented on ubiquitous and increasingly powerful mobile devices, which are simple to operate in the field and equipped with a large array of 3D spatial sensors. The project scenario employs available VOMs from the SAFARI database being quickly interpreted during fieldwork campaigns to maximise the information value of VOMs and fieldtrips alike. A high-fidelity geological field app for mobile devices is the practical outcome. The scientific focus is on developing and assessing novel techniques in geomatics, computer vision, computer graphics and visualisation, running on power- and performance-limited devices, allowing rapid and intuitive outcrop interpretations. The mobile device interpretations are further exported to be used for the MPS TI generation.

Outcrop photos are captured and coarsely geo-registered via mobile devices (e.g. tablets) during fieldwork. The photos are interpreted with respect to stratigraphy, sedimentology and their structural features. After establishing an image-to-geometry mapping, the interpretations can be directly projected and viewed in 3D during fieldwork on the tablet.

Veröffentlicht unter English, Geography, Geosciences, Work | Hinterlasse einen Kommentar

Announcement: Ph.D. successfully defended!

I have recently defended my doctoral degree at the University of Bergen on the subject of Visual Techniques for Geological Fieldwork Using Mobile Devices. The Ph.D. thesis is made available in the open access online publication portal of the university (BORA).

Kehl, C. “Visual Techniques for Geological Fieldwork Using Mobile Devices”, University of Bergen, Faculty of Mathematics and Natural Sciences, 2017

As a little, lasting mark of mine in Bergen, the department has put my poster on the ‘wall of fame’.

CKehl_PhDthesis_2017_poster

Poster – Ph.D. thesis – C. Kehl – Visual techniques for Geological Fieldwork Using Mobile Devices (2017)

Veröffentlicht unter English, Geosciences, Visual Computing, Work | Hinterlasse einen Kommentar

Announcement: PhD Defense at University of Bergen

I will defend my current Ph.D. for a Doctorate of Mathematics and Natural Sciences at the University of Bergen & Uni Research!

Title ? Visual Techniques for Geological Fieldwork Using Mobile Devices (VOM2MPS project)

When ? 6th October 2017 – 13:15 (Defense duration: ~2 hours)

Where ? Realfagbygget, Allégaten 41, 5008 Bergen, Auditorium 5

Portrait

Photographer: Daniel Francois Malan

 

 

 

 

 

 

 

 

Weiterlesen

Veröffentlicht unter English, Geosciences, Visual Computing, Work | Hinterlasse einen Kommentar

Creating a bibliography template in Word – example of the “Photogrammetric Records”

Despite the never-ending and dedicated efforts to convert the local geoscientific community to publish using LaTeX, thereare still a large sum of journals only accepting Microsoft Word-written papers. Though that may not be the most clever way of doing it, I understand that some people just end up with a computing device in their hands “by accient”, and that they want to get stuff done with minor learning efforts (by spending a large amount of time each repetion on the same task manually) – as the case with writing.

Weiterlesen

Veröffentlicht unter Allgemein | Hinterlasse einen Kommentar

3D NordOst 2015 – a researcher’s wrap-up

As in every discipline, the end of the year means also a time for research community gatherings, workshops and forth-so. For my part, this means to traditionally attend the 3D NordOst workshop on applied graphics, visualisation, processing and interaction of 3D data in Berlin/Adlershof. This venue is hosted for several years by the GFaI (Gemeinschaft zur Foerderung angewandter Informatik; engl.: Association for the Valorisation of Applied Informatics) in Berlin, taking place annually at the beginning of Decembre.

Weiterlesen

Veröffentlicht unter Allgemein, English, Graphics, Work | Hinterlasse einen Kommentar

International Supercomputing Conference 2015 – a researcher’s wrap-up

ISC 2015 (Franfurt a.M. / Germany) celebrated an anniversary to the conference. Although some standard contributors were absent this year due to political reasons, American and German attendance still peaked.

Weiterlesen

Veröffentlicht unter Allgemein, English, HPC, Work | Hinterlasse einen Kommentar

Numerical Methods in Visual Computing – Intro + Part I “Least Squares Optimization”

After several years as junior researcher in diverse disciplines and at diverse institutions working on Visual Methods, I came to the insight that it is not easy to focus one owns reading, research and personal development in Visual Computing. Numerical Methods are a key aspect in the majority of topics and techniques in the field. Although teaching staff members everywhere tend to criticise the mathematical level of interested student, many people underestimate one key aspect: There are so many numerical methods out, available and possibly important to the students – how can we expect beginners in the field to make a good reading selection ? The general will of beginners is to find good selection of methods to learn that covers the large extents and get them the furthest within the field.

Within this blog series, I’d like to propose a base selection of Numerical Methods that are of general nature and that cover a wide range of applications and application disciplines. The selection is meant hereby as a discussion basis for lecturers and interested people in the field of Visual Computing, but it may also serve as inspiration to young researchers to get in touch with some of the methods.

Weiterlesen

Veröffentlicht unter English, Visual Computing, Work | Verschlagwortet mit , , , | 2 Kommentare

Convolutional Neural Networks in PyLearn2 – Experiences with the Framework

I was working in recent days on experiments for large-scale Visual Object Classification (VOC) for Content-Based Image Retrieval (CBIR). Based on the commonly accepted high success rates of Convolutional Neural Networks (CNN), I tried to recreate Alex Krizhevsky‘s Convolutional Neural Network for my case study.

Weiterlesen

Veröffentlicht unter English, Programming, Work | Verschlagwortet mit , , , | Hinterlasse einen Kommentar

3D Scanning and Reconstruction of Large Scale Environments

One task in the field of virtual and augmented reality is the acquisition of 3D environmental scenes which can be used to integrate artificial three-dimensional objects. This is commonly done by using stereo camera systems for scanning the environment. The major drawback on these solutions is a high purchase cost for such systems. New technological solutions for three-dimensional scanning have emerged over recent years, such as Microsoft’s Kinect.
Therefore the aim in connection with the VR Laboratory of Aalborg University is to create 3D environmental scenes using low-cost depth scanning equipment. Connected with that task is the research on usability of low-cost equipment in terms of accuracy and noise susceptibility.
A variety of approaches has been analyzed and tested to solve each partial task ranging from calibration issues and pose tracking to 3D reconstruction, which are presented in this report. The final system uses a point based tracking approach realized with SURF feature extraction, a minimum-correlation feature matching and Gauss-Newton iteration-based transformation parameter extraction. The memorylimiting task of large scale point cloud storage is solved with a dynamically growing hybrid hash map. A per-frame image-based reconstruction algorithm is used for in-place augmented reality. Full scan reconstruction is done using a parallelized version of Marching Cubes.
As a result the current implementation does not meet needed real-time requirements, leaving certain space for SIMD and MIMD parallelization optimizations.
The system is limited by high noise susceptibility and the need for highly accurate feature points, in particular in point feature-poor environments. These limitations are tolerable regarding the low cost of the system.

Veröffentlicht unter English, Graphics, Work | Hinterlasse einen Kommentar

Conformal, multi-material mesh generation from noisy labelled volumes – DeVIDE FE-Mesher / CVMesh

In orthopaedics, finite element models are used for simulating force distributions in hip prostheses and surrounding bone. Such simulations can predict fracture risk or implant instability.
Patient-specific finite element modelling requires robust methods for converting volumetric scan data such as CT or MRI to tetrahedral meshes. Almost all biomechanical models contain multiple materials with vastly different physical properties that are typically organized in complicated geometric configurations. Extracting precise geometric models of the interfaces between these materials is important both for visualization and for realistic physically-based simulations.
There are few openly available algorithms that can solve the challenges of volumetric multi-material meshing of complex shapes. Several promising meshing approaches are described in recent literature. These algorithms are taken as a starting point for further analysis and research by the student. Algorithms are implemented on the open-source DeVIDE platform where possible, and optimized both in terms of selection of parameter and speed. The applicant should compare his algorithmic results with comparative software solutions from academics and industry.

Related Publication:

Kehl, C., Malan, D. F., & Eisemann, E. (2012). Conformal multi-material mesh generation from labelled medical volumes. In 3D NordOst workshop (Vol. 15, pp. 47-56). Gesellschaft zur Foerderung der angewandten Informatik (GFaI). article

Veröffentlicht unter English, Graphics, Work | Hinterlasse einen Kommentar