Enhancement of Hidden Cultural Heritage: The Submerged Forest of Lake Tovel

Hello everyone,

I would have liked to write this post long ago, but I couldn't find the time. As many of you know, Arc-Team has been involved in the study of the submerged forest of Lake Tovel since 2005 (a project started by prof. Tiziano Camagna). For years, especially since we started using 360° cameras, I have been thinking about how to make this hidden cultural heritage accessible even to those who cannot dive (the lake is in a natural park and special permits are required for diving, usually issued for scientific research purposes). The main problem with 360° cameras for underwater shooting was the lack of custom waterproof housings. I had even reached the point of designing one, but fortunately, the company Isnta360 began producing underwater waterproof housings for its models. Finally, a couple of years ago, we had the opportunity to test this technology underwater, thanks to a commission received from APT Val di Non. After some dives, we managed to obtain a good 360° video of the submerged forest. For this first step, the main difficulties were finding optimal conditions for diving. In fact, there are many variables that can influence visibility in mountain lakes. Firstly, there is the weather factor: recent precipitation, even of low intensity, can stir up the upper layer of silt deposited on the lake bottom, which thus goes into suspension and adds to the clay (practically always dissolved in the water), significantly lowering underwater visibility. Secondly, the seasonal aspect should also be considered: cold seasons generally guarantee better visibility, but obviously in winter the lake freezes and the access road is covered with a thick layer of snow. Finally, the diving route should be evaluated. Following some tests with our Insta360 One X 2, we saw that, at Lake Tovel (with good visibility conditions), we should not descend below 20 - 22 meters, otherwise the light in the shots tended excessively towards the blue-green hue and there were no underwater amber filters yet to correct this issue in this type of cameras.

However, in the end, we managed to obtain a good 360° underwater video, which, adjusted through the open-source software KDEnlive, was finally edited and uploaded to YouTube at this address.

 

Generally, YouTube automatically recognizes this type of footage, but perhaps because it was edited in KDEnlive, this time it didn't work, so I had to inject the 360 metadata into the mp4 format. There are various options for doing this, including the open-source software spatial-media. Once the issue was fixed, YouTube (and other video platforms) was able to "understand" that it was a 360 video and uploaded it accordingly (also activating the option of stereo viewing, in case you intend to enjoy the video through Virtual Reality headsets).

From the same footage, I was also able to extract some 360° panoramas of various Points Of Interest (POIs), in order to create, through the open-source software Marzipano, a specific App, which was then uploaded to the APT Val di Non website at this link. Below, you can see a screenshot of the App.

I hope the post is useful. Have a nice day!

ArcheoFOSS 2023, 17th edition: Torino 2023

Hello everyone.

As you may know, we at Arc-Team are particularly connected to the ArcheoFOSS workshop, a meeting on archaeology and open-source software that we helped found in 2006. Last December, the seventeenth edition took place in Turin (Piedmont, Italy) and, after a couple of years of absence, we were able to participate again. Our presentation was titled "Digital twins of archaeological finds. Open source technologies applied to 3D scanning. Methodologies, limits, and results" and dealt, among other things, with a topic that is very dear to me: how to carry out good 3D archaeological documentations (with open-source software) of artifacts with "difficult" surfaces, such as gold or fur.

Today, I took the time to upload our presentation, of which you can see the initial slide below. You can find our presentation at this address.

 

 

Below is a brief summary of the topics covered slide by slide:

SLIDE 1

Simply the title.

SLIDE 2

The institutions involved in the presented projects: MArTA (Museo Archeologico Nazionale di Taranto), Südtiroler Archäologiemuseum - Museo Archeologico dell'Alto Adige, UMST (Unità di Missione Strategica della Provincia Autonoma di Trento), and, of course, Arc-Team.

SLIDE 3

MArTA, perhaps the most important museum in Magna Graecia, famous for its collection of gold items.

SLIDE 4

Südtiroler Archäologiemuseum - Museo Archeologico dell'Alto Adige, famous for hosting the mummy of Ötzi (the Similaun Man).

SLIDE 5

The Autonomous Province of Trento, which, with its offices, is responsible for the conservation and enhancement of Cultural Heritage in Trentino (Italy).

SLIDE 6

Arc-Team: a company specialized in various archaeological sectors, including FFR (Forensic Facial Reconstruction), Field Archaeology, Digital Archaeology, Mission Abroad, Archaeorobotics, Extreme Archaeology, etc...

SLIDE 7

An example of the wide range of artifacts stored at MArTA.

SLIDE 8

An example of the type of artifacts associated with the Similaun Man's mummy (mostly made of natural materials such as wood, leather, fur, etc.).

SLIDE 9

An example of the various types of archaeological projects carried out by Arc-Team in the Autonomous Province of Trentino - Alto Adige/Südtirol.

SLIDE 10

A gallery concerning issues related to 3D documentation of artifacts in the mentioned contexts. Main issues include the size (from very large to very small) of artifacts; logistics (with non-dismountable exhibits, non-detachable display cases, old museum contexts, crowded visitor rooms, etc.); the complexity of some objects (which may require documenting both the internal and external surfaces in 3D, even in the case of very small artifacts); the variability of the characteristics of different surfaces (including extremely reflective ones, such as gold, silver, and sometimes bronze; semi-transparent ones, such as glass; those that absorb light, such as skin and certain types of wood; and very particular ones, such as fur).

SLIDE 11

The solution that Arc-Team prefers for 3D archaeological documentation: SfM (Structure from Motion). In the photo, Alessandro Bezzi, Luca Bezzi, and Kalus Kerkow in front of the TOPOI Excellence Cluster lion in Berlin (Germany), documented in 3D in 2009 with the open-source PPT (Python Photogrammetry Toolbox) software developed by the Frenchman Pierre Moulon.

SLIDE 12

The workflow PPT (now obsolete), which relied on the open-source Bundler and PMVS software.

SLIDE 13

The level of detail/accuracy achieved with PPT in 2009.

SLIDE 14

The current situation in the panorama of open-source SfM software. Among the various options: OpenDroneMap (excellent for archaeological excavation and aerial documentation), openMVG, Regard3D, MicMac (excellent for aerial documentation), orthoBlender (excellent for FFR and any project involving a 3D modeling step), Colmap, openMVS, MeshRoom (excellent in creating and managing raster textures).

SLIDE 15

A slide to show the versatility of a 3D approach based on SfM, capable of operating in extreme contexts (Underwater Archaeology, Aerial Archaeology, Glacial Archaeology, High Mountain Archaeology, Speleoarchaeology), as well as in logistically simpler contexts (Field Archaeology, museums), based on very simple equipment such as a camera, easily usable even during missions abroad.

SLIDE 16

A video of the 3D documentation of a speleoarchaeological context: the Grotta del Teschio, investigated by Arc-Team at the end of 2023 on Monte Stivo (southern Trentino, Italy).

SLIDE 17

Issues of 3D documentation of artifacts related to light: problems due to ambient light and problems due to reflected light.

SLIDE 18

A typical example of 3D documentation problems related to ambient light: underwater documentation (solvable with pre-processing photos through the open-source software ImageMagick). Specifically, the photo shows the example of the Barca dei Diavoli (which I have already talked about in various posts here on ATOR in the past).

SLIDE 19

The 3D reconstruction and vector drawing (from GIS) of the Barca dei Diavoli.

SLIDE 20

Another example of problems arising from ambient light, this time inside a museum (MArta): presence of poorly lit areas; artificial light spots; ambient light (sun) divided into three different entry points from the windows of the wall; presence of large reflective surfaces.

SLIDE 21

Solutions to eliminate or mitigate problems due to ambient light inside museums: dark or reflective panels, special flashes (ring-type), etc.

SLIDE 22

A video showing the example of 3D documentation of a marble head not removable from its museum location. Documentation is carried out through SfM with the help of a ring flash and two LED panels for uniform lighting.

SLIDE 23

Another example of correcting ambient light in a museum (in cases where small objects removable from display locations are documented), through the use of simple photographic boxes.

SLIDE 24

Problems with reflected light and the dynamics of light on reflective surfaces. The first image shows Blender's mascot (Suzanne) in gold, as it would appear in reality, with both diffuse and specular reflection light. The two derived images show, on the left, Suzanne with only diffuse light, and on the right, Suzanne with only specular reflection light. This slide serves to show the technique normally used to document objects with a very reflective surface through SfM: an attempt is made to obtain only photographs with diffuse light (without specular reflection light), from which 3D is then extracted.

SLIDE 25

In some cases, such as bronze, to obtain reflection-free photos, it is sufficient to use a polarizing filter on the camera (and rotate it according to the direction of the reflection). In the example shown in the photo, a bronze helmet kept at MArTA.

SLIDE 26

The 3D result of the helmet from the previous slide.

SLIDE 27

Another example of a complex 3D from MArTA: a semitransparent glass cinerary urn with the amphora in which was housed.

SLIDE 28

An example of 3D with reflected light from organic materials (wood, leather strings): Ötzi's axe.

SLIDE 29

The technique used by Arc-Team at MArTA for the most complex objects due to the high level of light reflection: gold items. The image reproduces the arrangement of a darkroom in the server room, with the polarization (via a film) of the only light source and the use of a second polarizing filter on the camera.

SLIDE 30

An image illustrating the removal of reflection from a gold crown.

SLIDE 31

A video showing the 3D model obtained from one of MArTA's most complex gold crowns (due to the complexity of the decoration).

SLIDE 32

The concept of Archaeological Tolerance to be applied to the level of precision and accuracy intended to be achieved in 3D surveys (both of artifacts and excavation contexts).

SLIDE 33

An example of a special case: the 3D documentation (for FFR) of the face of the mummy of S. Caterina Fieschi Adorno (from Genoa).

SLIDE 34

The difficulty was due to the fact that the crystal coffin in which the saint's body is preserved could not be opened. For this reason, Arc-Team documented the face by placing the camera directly on the glass slabs of the coffin. Obviously, with such forced shooting points and the low quality of the photos, due to the glass panels, a 3D was obtained that was not sufficient for the continuation of the work (the identification of restoration interventions, on the nose and mouth) carried out in the 1960s. Nevertheless, it was possible to increase the level of detail of the 3D documentation, thanks to Cicero Moraes (FFR expert at Arc-Team), who used a new technique in Blender (through the Map Displacement modifier), later integrated into orthoBlender.

SLIDE 35

This technique is very useful for increasing the detail of the docuemntations of particular surfaces such as fabrics or skin. The example reported here concerns a detail of Ötzi's leggings.

SLIDE 36

Another example of the Map Displacement Modifier, this time applied to Ötzi's bear fur hat. However, as seen, this modifier is unable to increase the detail of the docuemntation in areas where the fur has been preserved.

SLIDE 37

At the moment, the best method for rendering fur (at least in 3D videos) is the use of the new technique NERF (Neural Radiance Field), through AI and open-source software like NERFstudio.

SLIDE 38

The result of the 3D of Ötzi's hat in NERFstudio.

SLIDE 39

Of course, Archaeological Tolerance in 3D surveys of artifacts must also be calibrated based on the very purpose of digital twins. For example, Ötzi's artifacts were 3D documented for conservative purposes, while the skull of the Grotta del Teschio was documented as part of normal speleoarchaeological excavation operations (setting up a georeferenced coordinate system in the cave). The digital twins of MArTA's artifacts were created to be integrated into a webGIS, for scientific and educational purposes, developed by Arc-Team for the Taranto museum.

SLIDE 40

Conclusions: the only technique that allowed obtaining 3D (digital twins), acceptable according to Archaeological Tolerance, in all these cases was SfM. Any technique set up on different hardware has failed in one or more cases.

SLIDE 41

Among the most difficult objects to 3D document during the MArTA experience were the two monumental red-figure kraters (one attributed to the Painter of Carnea and the other to the Painter of the Birth of Dionysus). These artifacts, also due to their size, were not movable and were photographed through the glass of the display case, of course, taking care to eliminate ambient light as much as possible (also due to the large extremely reflective black varnish surfaces).

SLIDE 42

The 3D model shows the final result, while the photos on the right show the forced camera shots (green points).

SLIDE 43

Another of the most complex objects to document at MArTA was a small (working!) ceramic nymphaeum. For reasons of Archaeological Tolerance, it was extremely important, in this case, to document both the exterior and interior of the object, despite its small size. Only in this way could the functioning of the artifact be shown, which, through a small external tank, filtered water internally to then spout it from a small internal lion protome.

SLIDE 44

The internal documentation of the ceramic nymphaeum was possible only through SfM techniques, as no tool, outside of a small camera, could be inserted through the openings of the artifact (of very small dimensions). The image shown here is a 360-degree shot of the inside of the nymphaeum.

SLIDE 45

Thank you for your attention!

I hope that this presentation can be useful. Have a nice day!

Open Source AI and Archaeology: OCR and HTR

Hello everyone,
this is the first in a series of posts dedicated to Open Source AI (Artificial Intelligence) applied to archaeology. During my work, I have often found myself, lately, working with various AI tools.

Yesterday, for example, I was engaged in the recovery of old typewritten documentation regarding Lake Antholz/Anterselva, where Alessandro Bezzi and I will soon be involved in some underwater archaeological explorations (as part of a research project on the lakes of South Tyrol, in Northern Italy, entrusted to Arc-Team). To expedite my work, I used Open Source software called Transkribus, developed by the University of Innsbruck (Austria). In short, Transkribus is "an AI-powered platform for text recognition, transcription, and searching of historical documents – from any place, any time, and in any language."

The image below is an example of how easy it is to use Transkribus, thanks to the web interface developed by the University of Innsbruck. As you can see, I just had to upload the scan of the document I had available (unfortunately of low quality and with strong distortions on the left margin) to obtain a fairly faithful transcription in just a few seconds. Transcription errors are mainly due to deformations in the document (unfortunately, I didn't have anything better available). Nonetheless, even with my limited knowledge of German, I managed to correct the text and obtain a good translation into Italian (my native language) thanks to other AI tools (which I will discuss in other posts).

This post is primarily dedicated to OCR (Optical Character Recognition) and HTR (Handwritten Text Recognition), two branches of research that help us archaeologists (and obviously historians and archivists, more qualified in this) quickly read ancient documents. Transkribus can perform both OCR and HTR, and indeed I have often used it for reading cursive documents (for example, related to World War I or archaeological documentation from the late 19th to early 20th century, but these will be topics for other posts).

What I want to convey here is that AI is forcefully entering (for some time now) the world of archaeology (I will soon provide more examples in other posts), even in areas where it seems unlikely to happen, and there are often many Open Source options to use these "new" tools. For instance, there is Kraken, "a turn-key OCR system optimized for historical and non-Latin script material," released under the Apache 2.0 license, very useful, for example, for documents in Arabic or Greek (but not only). Kraken itself is linked to the "eScriptorium" platform, also developed with open-source code by various partners, including the University of Paris (France).

If you are interested in Open Source tools for OCR and HTR, I recommend starting with eScriptorium and from this page, where you can upload your first documents to run tests.

Have a nice day!

HumanOS, a French FLOSS for archaeoanthropology

Hello everyone.

For several years, Arc-Team has been working at the ancient monastery of S. Anna di Sopramonte near Trento (Trentino, Italy). The project, initially started as an archaeological excavation, has evolved over the years into a physical anthropology summer school, as a collaboration between the University of Padua (Italy) and Appalachian State University in North Carolina (USA). Beyond the actual archaeological excavation, various activities and lessons related to archaeology take place during the summer school, covering topics from computational archaeology to archaeoanthropology and geoarchaeology. Typically, my colleagues and I at Arc-Team are involved in both the technical direction of the excavation and the training of students, especially regarding excavation techniques, 3D documentation, and project management through GIS.

I must say that during the last campaign (July 2023), I realized how old I am and how much the new generations have improved in archaeological techniques: for the first time, more than teaching students, I learned from them. Specifically, I learned about the existence of HumanOS, the topic of this post. HumanOS is a French open-source software developed for the management of human remains during an archaeological excavation. Essentially, it is a DBMS for archaeoanthropology. Personally, I have only recently started using it (usually, I manage this data directly through GIS and WebGIS), and I must say that the software seems well-designed and well-developed, especially from an anthropological perspective. Probably, from an archaeological point of view, it might need some adjustments (at least to fit correctly into the workflow here in Italy), but at the moment, it appears to be an excellent option for managing anthropological excavation data, especially in projects that have reached the study of bones.

Below, you can see an image of a tomb (number 5) from the S. Anna excavation where the preservation level of the bones is recorded (which is perhaps the best aspect of managing archaeoanthropologic analyses through this software). I hope to post more updates on HumanOS soon. In the meantime, have a nice day!

 

 

Acknowledgement:

 

Many thanks to Nicol Rossetti, Viola Polastri, Vanessa Marras for introducing me to HumanOS