blog: Towards an Integrated 3D Archaeological Database- Keiss foreshore, Caithness
Research Summary: Keiss Foreshore, Caithness
With the aim of producing a complete 3-dimensional survey of an archaeological landscape, the relict landscape at Keiss foreshore was laser scanned to produce a highly accurate terrain model and integrated archaeological database. The Keiss foreshore area has been the subject of archaeological survey and excavation as part of the Caithness River of Stone project involving partners from the National Museums of Scotland, the University of Nottingham and AOC Archaeology Group, and constitutes a well preserved record of coastal settlement in NE Scotland with a range of buildings dating from the early Iron age to the medieval period represented.
A Mensi GS101 laser scanner was used to capture the foreshore area with a coverage of better then 0.1 net XY resolution, with high resolution scans of the 'broch villages also taken. The scans were related to Ordnance Survey referenced control, established using a total station and registered with Trimble's Realworks software. The registered point-cloud was then sampled using a topography-based algorithm and decimated to produce a GIS-compatible geodatabase. A 0.1m terrain raster was then produced using ArcGIS and contoured and hillshaded. Interpretive survey was then undertaken using DGPS (tied in with an on-site calibration to match the project coordinates) in order to record interpreted wall faces and other archaeological features. This was complemented with a narrative photographic record and description of the sites, referenced within the GIS using hyperlinked anchor points. Feature IDs, structural context numbers, photo numbers and view directions were recorded in the survey database in the field using pre-designed input forms running on the Trimble CU GPS controller. The result is a landscape survey of the foreshore incorporating both narrative and objective measured data.
In addition to the topographic survey, excavation at Whitegate broch was recorded using laser scanning. Context, sample and small find data were recorded with a Leica TCR705 total station and Penmap working on the same coordinate grid so that excavated records could be integrated with the overall survey of the foreshore. The laser scan and point data were then combined to produce an integrated 3-dimensional record of the excavation of one of the main sites on the foreshore.
The Keiss archaeological survey is an exercise in the production of an holistic archaeological record involving a range of data capture methods, and is one part of a programme of digital recording of archaeological fieldwork in Caithness, undertaken through the River of Stone Project.
| Attachment | Size |
|---|---|
| Picture1.jpg | 36.19 KB |
| image002.jpg | 25.79 KB |
| Picture2.jpg | 93.92 KB |
Your comments invited!
I'd be very interested to hear your thoughts on this project, any similar experiences etc....?
Looking forward to hearing from you!
Graeme
----------------
Dr Graeme Cavers
Research Associate
University of Nottingham / AOC Archaeology
Keiss foreshore survey
This excellent piece of work would undoubtedly have been enhanced by a high resolution multibeam sonar survey of the archaeological evidence in the sea. Point cloud data from underwater surveys can be of comparable resolution to laser surveys on land.
Martin Dean
ADUS
University of St Andrews
x-msg://55/martin@adus.org.uk
x-msg://55/mld@st-andrews.ac.uk
x-msg://55/www.adus.org.uk
Re: Keiss foreshore survey
Hi Martin
Agreed- this would certainly enhance the foreshore landscape survey. Some initial maritime reconnaissance work has been undertaken as part of the project but this has yet to include remote sensing. Hopefully one day...!
----------------
Dr Graeme Cavers
Research Associate
University of Nottingham / AOC Archaeology
Scanning - Moving topolgy toward the Z domain
We have a similar terrestrial Laser Scanner (TLS) and have been using over the past two years for recoding buildings, monuments and excavations. The technology really speeds up the recording process with the main improvement over traditional hand drawing being the removes some most of the subjectivity and a massive increase in accuracy and resolution.
One of the problems we have found up till now, at least concerning the TLS of excavations is the transfer of true 3D data into a GIS framework so that the 3D data can assist the analysis and not justprovide a pretty picture. As present we have to flatten surface to digitise detail as GIS is only 2.5D and volumes aren’t respected. We have had some success using Geomagic which is a rapid prototyping software that copes with the complex surfacing of scan data. The GIS community needs to develop true 3D analysis tool and not just surface analysis tools. What we need is topology not just in XY but also in Z.
Hopefully as 3D capture technology becomes embedded as a major primary collection tool the software companies will develop the solutions we require.
Re: Scanning - Moving topology toward the Z domain
Hi Anthony,
I think you have hit the nail on the head... At the moment, the best you can aim for is a kind of GIS compatible version of the data: a detailed elevation model is great for terrain but doesn't really cut it for intra-site GIS on excavations. Geometry modelling of the complexity of archaeological material is always likely to be difficult and time-consuming too (and defeats the point of labour saving through scanning). Perhaps one day 3D Analyst will handle point cloud data and provide a solution, especially since scanning software seems to be coming to the conclusion that point-clouds are as good a 3D product as any secondarily-derived data. The challenge thereafter would be to intelligently layer and assign attributes to such large datasets, but this undoubtedly looks like the way forward for scanning and GIS.
G
----------------
Dr Graeme Cavers
Research Associate
University of Nottingham / AOC Archaeology
X, Y and Z
Five years ago I attempted to model ashfall from a Bronze Age volcanic eruption using a TIN. This involved creating a GIS-compatible database of the sort Graeme is describing (or at least I think/hope so), integrating and standardizing X,Y and Z data from a variety of published an unpublished sources. It was a clunky approach, but I simply couldn't think of a better way of doing it. It's therefore very heartening to see that such issues associated with 2.5D reconstructions are still there (and, indeed that there wasn't a horribly obvious trick that I was missing). It strikes me that one problem is that, historically, we have thought in terms of X and Y pretty successfully at a regional, or even intra-site, level. But in terms of Z, we only record good data in very localized instances, where there is an immediate need to document information about elevation, as well as location in the 'traditional map' sense. (That is a general point, not a commentary on the project under discussion, which, on the contrary, looks like the exception that proves this rule).
So... I completely agree with everything said by Anthony and Graeme. My question/s would be what kind of data standards (or standardization) would be needed for this 'topology of Z', and is there any hope of extending this principle to historic, i.e. secondary, excavation records, to better facilite their integration and digital representation?
Pointclouds in GIS
I have been experimenting with LP360, which looks like the best stab I have seen at integrating pointclouds to a GIS environment. It is designed for LiDAR, and there are lots of useful tools for processing terrain data. With a bit of work to edit asciis etc it is possible to get TLS data in too however, which allows you to work with pointclouds in a GIS environment, so you can layer your data and integrate shapefile datasets etc. It looks like it might be a very useful GIS extension for archaeology, and might answer some of the questions raised by this discussion thread, though there are a few limitations. Has anyone got experience with this or anything similar?
You can get a demo version from the website, which is: http://www.qcoherent.com/
Cheers,
G
----------------
Dr Graeme Cavers
Research Associate
University of Nottingham / AOC Archaeology
similar project in France
Hello,
we have made a very similar project in France with the Europen Aercheological Center of Bibracte (France, Burgundy). We have use a GS101, and digital photography. I thing we didn't go as far as you had, but it could be very interesting to do a complete 3D location of the archeological materieals found in the site.
If you could send me more informations, I would be pleased to send you back elements on our experience.
Best regards,
Ghislain Moret