Alex,

 

You've raised an interesting question.

 

In our experience, the best way to proceed very much depends on what you want to do with the 3D model once you've constructed it ... i.e. what's the scientific goal? The answer to this will have implications for the spatial precision and accuracy that you require in the model, and that is then likely to influence the best workflow to use. Perhaps you could tell us a bit more about what you want to achieve with the 3D model?

 

I share the enthusiasm that Terry Pavlis and others have expressed for the potential of structure-from-motion (sfm) photogrammetry in revolutionising outcrop geology. We have built dozens of models in this way, mostly for use in characterisation of fracture networks as input for DFN modelling, and the outputs can be really fantastic.

 

A word of caution though (before the hype overtakes the harsh reality of making these techniques as useful as possible) ... it's not always straightforward to construct outcrops with uniformly high spatial precision. And it's not necessarily clear when/where spatial precision is reduced, unless you have put in place good independent control during your survey. For some scientific purposes, that won't matter, but for others it may be critical.

 

We have analysed a range of different outcrops that we've recorded twice, once using sfm and again using a combination of two other independent survey methods (typically based on atmospheric corrected dGPS and lidar-based triangulation), and then compared the resultant virtual outcrops. Some of the sfm models used photos taken from UAVs, most are from ground-based photos. The results can be  great - but they are not always. We've learned not to assume that the photo-matching precision reported by the PG software (such as Agisoft Photoscan) equates directly to spatial accuracy in the final model. Just because the resultant outcrop looks nice, doesn’t mean that it is correctly oriented, scaled, or that it spatially matches the topography consistently across the outcrop surface.

 

In general, taking the photos of the outcrop is (usually) easy. Getting sfm software to combine these into some sort of 3D model is (often) easy easier ("just sit at your desk and press the buttons"). Rigorously ensuring that the 3D model is fit-for-purpose is IMHO a really critical step, but one that is easiest to ignore in the heat of the hype-wave.

 

So what? Well, I think that in order for us (i.e. the scientific community in general, and structural and tectonics geologists in particular) to be able to realise the potential of photogrammetry (and other rapidly evolving spatial methods) our appreciation of surveying principles and methods will need to improve - we need a better basis for understanding the quality of the spatial data from which new science will be derived.

A couple of concrete suggestions:

·         It would be great if journal editors and reviewers put pressure on authors to document the external survey control for all virtual outcrops used in papers (yep, not very exciting perhaps, but could be included as supplementary online material).

·         More papers published with quantitative comparisons of outcrops created using different technologies.

·         If they don’t already … undergrad classes should include sessions on surveying principles and techniques, 3D spatial errors/precision/ accuracy) etc. Maybe also basic photography principles (and practice using dSLR)?

 

 

A couple of papers that are really relevant:

 

James, M.R., Robson, S. 2012. Straightforward reconstruction of 3D surfaces and topography with a camera: Accuracy and geoscience application. Journal of Geophysical Research: Earth Surface. 117, F03017, doi: 10.1029/2011JF002289.

 

James, M.R., Robson, S. 2014. Mitigating systematic error in topographic models derived from UAV and ground-based image networks. Earth Surface Processes and Landforms, 39, 1413–1420, doi: 10.1002/esp.3609

 

 

Sorry Alex, if this has strayed a bit off-topic for you – but hopefully it may help save time & effort later down the line. What do you intend to do with your 3D outcrop model?

 

 

Richard

 

-------------------------------------------------

Dr. Richard Jones, Managing Director

Geospatial Research Ltd.

Office Suite 7, Harrison House,

Hawthorn Terrace

Durham DH1 4EL, UK

Office:     +44 (0)191 384 1759    

http://www.geospatial-research.com/

https://uk.linkedin.com/pub/richard-jones/19/ab1/799

 

 

-----Original Message-----
From: Tectonics & structural geology discussion list [mailto:[log in to unmask]] On Behalf Of Alex Brodeur-Grenier
Sent: 01 September 2015 16:08
To: [log in to unmask]
Subject: Cliff face photos in 3D space

 

Dear all,

 

I’m currently working on a 3D geological model of an open pit located in a polydeformed metamorphic terrain and I would like to import and georeference cliff face photos in 3D space. The photos have control points in X,Y and Z. Does anybody have experience on how to proceed for doing this? I would preferentially like to use the Surpac, Move or GeoModeller softwares, but, perhaps there is a better choice?

 

    Thank you for your help,

 

 

Alex Brodeur-Grenier

M.Sc.candidate

Polytechnique Montreal