View source

'''Please edit me using the edit tab and then just use the text box and the buttons at the top for links etc to correct/ change and make more sense of my notes!'''

This was based on real time notes of '''ArchCamp 2''' ('pron. 'Arcamp Deux') at CAA UK 2007 hosted at the University of Southampton on 27/01/07. These notes were made by Kayt. If anyone has images please upload them to the wiki!

----


== Intro ==

After the leisurely but pleasurely ArchCamp Un [http://www.antiquist.org/wiki/index.php?title=ArchCamp_Blog/Wiki], ArchCamp Deux was short and very very sweet. Amongst the CAA Survivors Club were:
* Leif Isaksen
* Stuart Eve
* Graeme Earl
* Michael Charno
* Kayt Armstrong
* Joanne Cook
* Eleftheria Paliou
* Michael Athanson
* Nic
* Matt Jones
* Hembo Pagi
* Bill
* Mercedes
* [http://www.antiquist.org/wiki/index.php?title=User:Cas Claire Louise Kemp]
* Evangelia Kappa
* Dave Wheatley
* Dave's Family :-)
- add/correct the names of any I've forgotten! -
 
We kicked off with a brief history of ArchCamp from Stu and Leif followed by a short discussion about setting dup ata-sharing webservices - a subject close to several of our hearts - but as time was pressing (a lot of trains needed to be caught at 2.00ish) we rattled quickly on to demoing stuff.

== SVG Mapping ==

The first demo was from Michael Charno from the ADS about an SVG/AJAX app for mapping tools that uses mouse overs etc to display and interact with mapping data.  It uses Ajax to dynamically interact with an XML file residing on the server which has attribute data.  As a user mouses over a 'tract' in the SVG image, a request is made to the XML file for the attribute data relating to that area.  The hope is that in the future, the tool will make requests to a more significant dataset (ie. the projects database) which has more useful information.

The tool was designed with the principles of Virtual Research Environments (VREs) in mind.  It's aim is to enable online collaboration and hopefully sharing of resources.  A paper in Internet Archaeology to come but the current application can be found [http://www-users.york.ac.uk/~mdc502/VRE/index.html here].  You need the SVG viewer from Adobe to view SVG images and currently IE is the only browser that works consistently.  There is a hack to get SVG images to work in Firefox but i need to get around to listing the instructions on my website.  Leif asked about it possibly doing 3D stuff but unfortunately SVG can only do 2D graphics.

== ARK ==

This seemed to work alongside with some things Stuart has been working on to do with Mapserver which might save some of the hard work!

His work involves making site databases for archaeologists (amongst other stuff). They got fed up of writing new databases each time to cope with varying recording strategies so they developed an 'open' datastructure where the archaeologists could pick whoch fields/data they wanted to record and have on thier context sheets etc, so that the recording strategy is led by the archaeologist, not the database.

Data then gets uploaded via a wiki-type interface via the web. It links photos and spatial data with the database- contexts sections and areas are digitised as sep, shapefiles that are put together by mapserver, which you can then examine with php and make spatial queries. It has filters as well to help navigate the data, which you can save, and also save time-specific outputs (i.e. this is what this query told me on 07.02.06) This also applies to the mapping data as it's designed for use on site so it's constantly updating.


== 3D Data Provenancing ==

Graeme then demonstrated something different and not as developed - he strongly feels we need to be able to apply attribute data to 3D models (metadata) to log model production and to discuss levels of certainty. Going against the opensource grain ;-) he was discussing using autodesk to do this, or ADO support available in 3D studio max. This allows a mini database of notes and data to be built up and attached to bits of drawings and so track your process of drawing - it can be queried and if you complete it as you go it will almost recall/ build a schema of how you built the model.


== Cool Websites ==

Leif, Graeme and Nic then showed us some websites which they think provide interesting services:

'''SIMILE from MIT''' [http://simile.mit.edu] is a website and project devoted to providing free tools and webservices for the Semantic Web. It has lots of stuff to pay with, more than could be demoed at ArchCamp - for example, '''Exhibit''' [http://simile.mit.edu/exhibit/] uses AJAX and a simple data file (in JSON format) to create interactive webpages. If it has spatial data it can also integrate Google maps, and temporal data can be displayed on a timeline. '''Babel''' [http://simile.mit.edu/babel/] is a further tool that lets you convert data into RDF, JSON and other SemWeb-related formats.

He also (just coz it was kewl) showed us the '''MIT lecture search''' [http://www.galaxy.csail.mit.edu/lectures] - they've built an engine that autotranscribes and indexes all the MIT lectures (past and present) so you can search the lectures by words people have actually spoken! They apprently intend to provide it as a free service so that you can upload your own audio data and have it transcribed. Uebercool.

'''The EPOCH 3D Webservice''' [http://homes.esat.kuleuven.be/~visit3d/webservice/html] builds 3D models from photos taken on normal cameras. You do need to have specific set up of distances and angles, but the technology required is very cheap. You upload images to the website and get models out of the other end.

'''CHI create Polynomial Texture Maps''' [http://c-h-i.org] that let you move a lightsource around images - it lets you change how an object is lit in a picture by some really funky maths, bringing out surface texture and shape in an amazingly lifelike way. It also lets you extract the geometry and build 3D models, and they are only about twice the size of a jpeg equivalent- much smaller than true 3D models. It will also do fun stuff for GIS azimuth data - terrain surfaces ! You do need a special rig for taking the images, however. It's sort of like how bullet time is filmed, using several lights and cameras. This is going to be used for all the finds at the Portus Project (link needed please!).

The MS Photosynthesis project appears to have morphed into '''MS Phototour''' [http://phototour.cs.washington.edu] There's a video of using it as a way of organising/searching through large photo collections, arranged spatially. The 1st view is of a point cloud, and you can then can drill down through the data. The photos are all georeferrenced to this common model. You can also transfer annotations from other datasets (e.g., Flickr), and it has non-photo mapping that it builds from the photos to let people navigate the images.


== Crossbones ==

Leif then demo'd the Crossbones application he has been working on for OA. Two sacrificial victims (thanks Mike & Stu) were shot in with a total station. Points are shot in order around the skeleton, going down the centre of the body then the left and finally the right side, in accordance with a simple schema. If a body part - e.g., arm - is missing, the relevant point ids are simply missed out. We had some problems getting the data out of the TS(!) so Leif then used a real dataset to show how the output of crossbones works in a 3D DXF visualisation progam (in this case ArcScene) actually works. In an ArchCamp exclusive, we even rigged it up to a passive stereo screen so that we could all explore a mass burial pit in ''actual'' 3D. Scary. Once the we've got the ArchCamp TS data Leif will convert it and stick it on the wiki for folks to look at and compare to The Southampton Chainsaw Massacre ;-)


== Bullet Tracing ==

Mike did a demonstration of his trajectory of bullets stuff that he is doing for his PhD.
It outputs scripts to generate a 3D model. He is trying to associate bullets with firing points and reconstruct sequences of events. Traditionally, viewsheds are used for this kind of work and it doesn't work very well.  - (please someone put in more information on all this!)

He also gave us our ArchCamp Word Of The Day - '''Thematise''' (''v.'') to organize one's dataset by theme.


== Wrap up ==

At 1.30 we said our tearful farewells to those catching trains and deCamped to The Crown for the traditional Beer and Burger. A massive thank you to the Southampton Archaeological Computing Lab for letting us play with their cool toys and to everyone who participated and made it such a fun (not to mention stimulating) event. 

Who knows what ArchCamp Trois may bring...

PS. Some pics about the ArchCamp and CAA UK 2007 can be found from Hembo's Flickr page [http://www.flickr.com/photos/hembo/sets/72157594505299364/], and at Kayt's [http://www.flickr.com/photos/kayt_lilith/sets/72157594573441583/]

[[Category:ArchCamp]]