From the beginning I must say that this post is not a guide. I recently published a paper on this topic, and I believe it is more than helpful for anyone in the position of digitizing translucid or highly specular objects, especially in the constraints of Cultural Heritage preventive conservation rules. So this post is a discussion starter and a kind of an invitation to read this paper.
The most spectacular objects to digitize are, no doubt, museum and archaeological artifacts. At least that’s how I see it. A perfect detailed reproduction of an intricate old piece of armor or domestic tools that nobody uses now, is truly fascinating to gasp at.
The possibility of remote manipulation and close-up visualization of virtual replicas of real historical artifacts without really touching them, is one of this age’s technological achievements (and wonders, why not?).
As I explained in other posts, the 3D digitization of Cultural Heritage assets is nowadays of critical importance. These Cultural Heritage assets come in different shapes and sizes: from minuscule fragments or objects to large architectonic or monumental ensembles. Today we can tackle with great success even at low costs most of the digitization cases in Cultural Heritage. With a very few exceptions which poses some difficulties and requires special attention.
These exceptions are made by a few types of materials that conflicts the nature of the available digitization methods. I another post I detailed the main methods for 3D digitization in Cultural Heritage and how and why they are optically based. This optical nature of these methods prevents them from being efficient with objects that have materials with special optical properties: transparency, translucency, specularity (strong reflections) or surfaces that changes color depending on the observation angle.
Translucid and specular materials
Translucid materials allow light to pass through them but diffuses it. When light interacts with a translucid body, it propagates through the object material, is scattered, and then it comes out through the incidence surface or it continues to propagate through the whole body. Direct illumination of these objects, due to light sub-surface scattering, results in poor detail detection by an optical sensor. For this reason, laser scanning systems have difficulties in correctly recording the 3D coordinates of the shape because the laser beam is not reflected back right from the surface.
From this class of materials, cultural heritage objects usually contains old glass, wax, marble, organic materials, special properties glass etc.
Specular materials have high reflectivity. We simply call them shiny. Shiny surfaces, depending on their geometry, have specular reflections on certain angles of observation (Fig. 1 C). These reflections are characterized, from the observation point, by a lack of detail and information, due to the light concentration. That’s why the detector cannot capture shape or color. This is a problem for both 3D laser scanning and photogrammetry.
From this class of materials, cultural heritage objects usually have metals, glazed surfaces, glass, marble etc.
In the industry there are successful methods to record the shape of objects with such materials. There are special powders and sprays that can be applied on the surface to deal with both translucency/transparency and specularity. The problem with Cultural Heritage objects is that this approach is not allowed. Not recommended is too kind of a statement. Most of the times this powder coating is difficult to remove without using force (brushes, water or air) which is putting the artifact at risk. Another aspect is the chromatic properties of the surface that cancelled when using a coating. Both surface morphology and chromatic properties are critical in 3D digitization of Cultural Heritage objects in regard with the documentation of their conservation state.
Are there any solutions?
There are several approaches, some more complicated than others. In my published paper “3D Digitization of translucid materials in Cultural Heritage objects: a comparative study between laser scanning and photogrammetry.” I approached this topic by fully detailing the physical properties of these materials and exposing and explaining the limits of current 3D digitization methods in this regard.
In a dedicated section I synthesized the state of the art and all the important studies made on this subject. From laser induced flourescence scanning or UV sensitive liquid submerging to UV lighting photogrammetry. These are great results and some of the approaches can be used, if you have the required equipment. Nevertheless, there is a lot to learn from all these studies.
The paper focuses on the comparison of two main methods of 3D digitization: 3D laser scanning and photogrammetry. While laser scanning is straight forward, photogrammetry allowed several tweaks like advantageous light positioning and using of linear and circular polarizing filters (cross-polarization). These methods are fully described, as well as the data acquisition parameters and conditions. Likewise the processing steps are fully detailed. All these so that the reader can make a full image and context of the measurements and so that he can extract his own conclusions.
The subjects were a candle wax sculpture and a jade statuette. Both qualifying for the topic criteria: translucid, highly detailed and also one of them, the jade statuette, presented high specular reflections.
The results are compared both visually but also mathematically. I used CloudCompare for direct comparison, which beautifully mapped on the surface of the two subjects the difference between the two methods.
The gallery below display several images with results and comparisons but also data acquisition setups highlighting the light scattering phenomenon during the laser scanning. As some images are cropped in the tiled gallery for a better fit, please click on each of them to see them at full format.
I won’t reproduce the paper here, but I totally recommend you to check it out on Researchgate. It is set on private but I will grant access for anyone who wants to read it.
Citing details: L.M. Angheluță, R. Rădvan, 3D digitization of translucid materials in cultural heritage objects: a comparative study between laser scanning and photogrammetry, Romanian Journal of Physics, 65, 906, 2020
Thank you for bearing with me, see you next time!
Have a great week! Cheers!