Jump to content
ArcheoAndy

Making a Static Dome

Recommended Posts

I am coming at this as without much knowledge or expertise.  So if this is a dumb idea, please let me know.  I have seen the results of this process and would like to give it a try.  My interest is in documenting small items only.  Generally less than 5 cm max length and generally consistent is size.  In order to make this a valuable comparable tool, I would need to make a large database of objects.  The idea of taking manual light distance measurements on hundreds to thousands of items is daunting enough to make this process unworkable.  My thought was to literally take a hard plastic bowl of sperical dimentions and cut holes in it at the appropriate light locations (leaving enough structure to make it remain inflexible).  I would then mount the camera through a hole at the top of the plastic dome directly over the subject and move the light source from position to position taking photographs without touching the dome.  The goal is to find a systematic way to make data collection faster on consistent size objects in controlled environments.  I can do this in situations where all other light is off, but I am not sure if there would be interference from reflected light within the dome.  Also, for focusing on small items only, can I use a smaller light source (preferably continuous)?

 

Any thoughts on making the process more systematic in controlled environments would be helpful.

 

Thanks,

 

Andy

  • Like 1

Share this post


Link to post
Share on other sites

Another example of a dome is the Portable Light Dome (PLD) designed at the University of Leuven (which I found from Graeme Earl's very useful RTI timeline).  There's a fairly detailed paper describing an earlier version of the PLD design, construction, and use in the bibliography at the PLD link, but I'll post it here.  Their current system uses 260 LED lights, which seems like a lot for simple RTIs and would lead to large storage requirements.  They have their own unique file format for building and viewing RTIs.  I'm sure the system could be easily used to build RTIs using CHI's software. 

 

With such a large number of LED positions, I think it would be interesting to modify the dome to capture multispectral RTIs with LEDs in groups of different wavebands, using a camera modified to allow IR, UV, and visible imaging, or ideally, a camera with a monochrome sensor (without a Bayer filter).

 

I'd like to see more examples of detailed dome designs published like this, and a compilation of resources for LEDs and materials to build domes would be really useful.  The folks at KU Leuven have said they want to make their dome available to others affordably, but I wasn't able to get any responses to my two e-mails about 6 months ago.

Share this post


Link to post
Share on other sites

Hi ! 

 

I discover RTI 2 years ago, and I tried applying this technique occasionnally in the musuem where I work in France. (sorry in advance for my english)

 

I use the highligt method with the kit and it works great. But for A project of referencing a large bundle of paleolithic graves stones, I would like to build a static dome (diameter 60 cm, 48 lights, capture session automated)

 

I found an electronic school, who will work on the software part for dialog between DSLR and lights.

My part involve designing the dome itself as I'll be final user. I already find a lot of solutions at HP or here.

 

But I'm looking for some feedback and have some technical questions and that's why I'm waking up this post to centralize informations and return of experience.

 

Advantage and inconvenient of led or incandescent for lightning ? 

 

Solutions tried for adjusting lights power for having the right exposition at different aperture ? a potentiometer ?

 

What thing we must pay attention for?

 

 

Thanks in advance for your answers.

David

Share this post


Link to post
Share on other sites

Incandescent  has more output in IR than a standard LED, but LED efficiency is much higher, so if you'll primarily be working in the visible LED  may be a better choice. Not sure what you mean by "right exposition at different aperture". If you mean exposure, you want all the lights to have the same intensity, and take all the photographs at a constant aperture and shutter speed to maintain the same exposure parameters for every photo.

Share this post


Link to post
Share on other sites

Hi everyone,

 

I'm new to the field of Reflectance Transformation Imaging so this might be a very basic question but: how does one synchronize all the different LED's in a static dome to the DSLR camera shutter?
 

Thanks!

Share this post


Link to post
Share on other sites

Typically done with a custom electronic control system that turns each LED on and off in sequence, triggering the camera shutter in sync through a remote control. You could also do it manually, powering up the LED and firing the shutter by hand.

Share this post


Link to post
Share on other sites

Thanks, Leszekp!

 

I was also wondering why the working field within the dome should be between 65 and 15 degrees? What would happen if I position lights outside of this range?

Share this post


Link to post
Share on other sites

Good question, and I don't know the answer. I've seen domes with higher angles, and my dome has a row of lights slightly outside that range (72 degrees) without any apparent ill effects. Perhaps this is just for Highlight-RTI? Maybe Carla or Marlin can chime in with an explanation.

Share this post


Link to post
Share on other sites

Thanks for the answer leszekp.

 

I don't know why I only think high speed and not long time exposure. It was the end of the year and and I really needed to take vacations.

 

I noted for the leds. Effectively we’ll work only with visible lights as our objects are more monochromatic surfaces.

 

Did you know if IR lights always give more relief detail regardless of the subject ? or it is just useful when you work with something colored (painting, ink traces) ?

Share this post


Link to post
Share on other sites

Since the camera is fixed/static, you don't have to worry too much about blurring from long exposures.

 

I have no experience with IR RTI. Here's a link to a recent presentation by Kathryn Piquette, where she used IR-RTI on Herculaneum papyri:

 

http://www.slideshare.net/UCLDH/piquette-ucldh-seminar20141203forweb

 

I think the University Of Southampton is also working with IR-RTI, but couldn't find a reference.

Share this post


Link to post
Share on other sites

Hi folks,

 

Sorry I have been absent from this thread.  Just got a bit tied up with other matters.

 

First - we recommend the 15 - 65 degrees because tests have shown that is the most effective area to work within.  If you have lights at higher angles, you may have the subject be too bright.  Similarly, with lower angles the subject may be too dark.  The software will work, but will you have a good set of images that are within the dynamic range of the camera?  This is also a good time to remind folks that for RTI the algorithms assume that everything in the environment is the same except the light position.  So the camera should be set to manual, the focus should be manual and the lights need to be of consistent power.  When you process the images, any modification, such as white balance and exposure compensation should be applied equally across the image set.  Just to be clear, the software will not complain and it will produce a result, however that result may not be correct in terms of the surface normal data, and may exhibit other problems.

 

The domes (light arrays) we build at CHI use continuous lights, and are controlled via different controller boards (depending on when we built the unit).  We communicate with the camera using a USB cable, and also Canon makes an SDK for folks writing software where you want to communicate with the camera.  There are other approaches out there as well.

 

Regarding the IR question, IR  (or UV for that matter) are useful in some situations with some kinds of material.  In the same way that shooting an IR or UV image might be helpful.  If you aren't seeing anything different in the IR image, then you won't get more information by doing an IR RTI.  IR is usually used because you want to examine some details below the outer surface that will reflect an IR band where that IR band passes through the outer layer.  For example with certain kinds of paintings.  The topic of multi-spectral imaging is a large one, and if folks want to discuss it further, I think maybe we should start a new topic in one of the other forums.  I'll suffice it to say that IR does not always show more or different details, and also there are many possible wavelength bands that are "IR"  What you can collect s dependent on the sensor you have and what it is able to collect, usually also constrained by either the light source, or by filtering the camera lens.

 

Carla

Share this post


Link to post
Share on other sites

Also - here is a really interesting multi-spectral and RTI project:

 

http://palimpsest.stmarytx.edu/integrating/

 

Note that all of the data is available, if you want to try some of this yourself.  Thanks to Todd R. Hanneken not only for doing the project, but for making everything availabl to others!

 

Carla

Share this post


Link to post
Share on other sites

Here's an interesting LED-array dome that Tim Zaman, a robotics Ph.D. student at the University of Delft, demonstrated at the receont 2+3D Photography conference at the Rijksmuseum in Amsterdam (the Youtube video is in Dutch):  https://youtu.be/Q8Gc7evFmnI

 

The dome provides a very entertaining light show!  As I recall, Mr. Zaman mentioned that he built the dome using addressable LED circuit boards that make up the structure of the dome as well as the lighting system, and he estimated the cost to be ~€100.  This limits the size of the dome, but I'm sure similar designs could be made larger (possibly requiring more powerful LEDs).  If anyone here speaks Dutch, perhaps they can translate or paraphrase the video for us.

 

The paper Carla linked to above, "Integrating Spectral and Reflectance Transformation Imaging...," suggests that capturing RTIs in many wavebands doesn't necessarily provide more information than you can get by capturing a set of spectral images without RTI and integrating the results with visible-range RTI.  In other words, the benefits of having multiple-wavelength LEDs in a dome might not be worth the additional expense and time to process so much data. 

 

However, IR-RTI has been used effectively to study papyri, as leszekp mentions above, and here is a link to a Southampton case study by Eleni Kotoula, a Ph.D. student there (I believe she has since graduated from the program):  http://acrg.soton.ac.uk/blog/3974/

 

She also studied UV-reflectance and UV-induced visible fluorescence RTIs, here:  http://www.academia.edu/11625431/Ultraviolet_RTI

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×