Damien Vurpillot (damien.archeo@gmail.com)

Tours, 11/09/2017

  tech enthusiast

  Where I come from

  Travels

  Travels

  Agenda

An open-source JavaScript library for world-class 3D globes and maps

https://cesiumjs.org/

- Open-source

- Can be associated with databases (SQL and NoSQL)

- Heterogeneous data (mesh, point cloud, raster, vector, etc.)

- Efficient file formats (glTF, entwine indexing for point clouds, etc.)

- Active development

​

Multi-resolution mesh example.

Open source photogrammetry software

http://micmac.ensg.eu/index.php/Presentation

Historical photogrammetry example, mapping the evolution of Besançon between 1940 and 1956.

It could be a big improvement over PattyVis for such an ambitious project.

chicago.vision is good example of "Big Data" management capabilities with Cesium.

http://chicago.vision/

A digital workspace to visualize and annotate multimodal and multiscalar 3D data, for archaeology and cultural heritage.

Mixing a professionnal social network interface with highly efficient web visualization tools.

.Point Cloud (PCL)

Test: Spatial footprint, density, homegeneity

Cleaning: SOR filter, density threshold (relative to the object and its size)

Indexing

Massive (>200M) = Entwine

else .laz or .pnts

big files

performance

light files

less efficient

Mesh

multiresolution glTF

.Multiscalar and multimodal

1. Landscape = PCL (Cesium)

2a. Site (excavation) = Mesh (Cesium) + PCL (Potree@metrology windows)

2b. Building = Mesh (Cesium) + PCL (Potree@metrology windows)

3. Dynamic layers (excavation step, etc.) = PCL (Potree)

4. Artefacts/Objects = Mesh (Aframe? => glTF viewer)

Each element from 2 to 4 has a queryable spatial footprint (polygon)

Front End

Entwine,​

.laz, .pnts, etc.

.gltf (3D tiles)

spatial data

(json, kml, czml)

Database

Metrology (isolation) =

Potree

Visualization (isolation) =

AFRAME

2

Going BIM !

In vulgar language:

- it can match fragments

- it can predict missing geometry from broken objects

- it can perform surface analysis

PRESIOUS will research and develop innovative methods and technologies to augment the geometric data of digital cultural heritage objects by predictive generative processes guided by geometry processing, analysis and shape matching methods for 3D objects.

http://www.presious.eu/

(not precious.)

Virtual Anastylosis ?

A process using range based and image based modeling digital techniques, makes a 3D model which reproduces a hypothetical copy of the original model, assembling surveyed fragments, present in the excavation site, with elements philologically reconstructed on the basis of historical knowledge and documentation.

Canciani, M., Falcolini, C., Buonfiglio, M., Pergola, S., Saccone, M., Mammì, B., & Romito, G. (2013). A method for virtual anastylosis: The case of the Arch of Titus at the Circus Maximus in Rome. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences II-5 W, 1, 61-66.

Virtual Anastylosis ?

.Photogrammetry = heterogeneous meshes with too many faces

1st decimation = averaging vertex density over the Via Appia artefacts to reach a relative homogeneity and a balance between accuracy/face count.

Presious pipeline relies on matching fractured surfaces and, optionally, enhance the matching by using prominent linear or curved structures on the intact regions.

A careful decimation can "smooth" surface noise and make relevant features more  prominent.

Presious perfomance can get very bad very quickly with unoptimized meshes (memory usage above the sky).

It could also be a good idea to perform the subsampling directly from point clouds, since the decimation process from meshes may harm the spatial integrity of the object (even with specific rules enforced to "preserve" it).

Virtual Anastylosis ?

1st decimation

An initial spatial segmentation with a subset of artefacts, more likely to match each other, could help to perform the analysis on such a huge dataset (>2500 artefacts).

We can then iterate with non matching artefacts.

.Serious Gaming

2nd decimation = to make meshes "game ready"

Baking Physics Simulation = baking the results from Presious before rendering (real time physic simulation is killing the performances).

Starting Physic Interactions = only when artefacts bounding box are close to each other, meaning people are acting on said artefacts (magnetic link).

Virtual Anastylosis ?

.Serious Gaming

"Portal" game concept with multiple stages of "increasing" difficulty so that you don't load the entire Via Appia.

Mixing Presious matched artefacts with non matching artefacts.

Asking people to match artefacts by themselves in a more or less guided way (color difference between pre-matched and non-matched artefacts, etc.).

Visual representation of the "goodness of fit" (color, heatmap?).

Log results of the interaction to enhance our matching database.

Ludic way of performing a boring task.

Could also benefit a lot from the added value of Virtual Reality: space/volumetric perception and natural interactions with the controllers.

Virtual Reality ?

Booming growth of a new ecosystem surrounding the virtual reality industry

Added value: Immersion and enhanced volumetric perception

Added value: More natural interactions with 3D data

It can be a great tool but it is limited to circumstantial application and comes with technical and physiological constraints.

Augmented Reality ?

Has been around for a while, the big innovation being augmented reality glasses.

Added value: "Augmenting" reality with a new layer of 3D information.

Added value: More natural interactions with 3D data

The technology does not seem to be here "yet". The field of view is really narrow (30°×17.5°) and the computing power of wireless devices is limited.

It's enhancing one’s current perception of reality, whereas in contrast, Virtual Reality replaces the real world with a simulated one.

Hololens field of view.

Still, I have no doubt it will become an indispensable professional tool in a few years.

Yes, you can look at Cesium in Virtual Reality with your smartphone and a cardboard.

For the moment, it's more a gimmick than anything.

https://photogrammetrie-mshe.univ-fcomte.fr/site/Chambord/Apps/Sandcastle/gallery/chambord_m.html

Aframe (from Mozilla) is on its way to become the next VR standard for the web (https://aframe.io/)

Virtual Reality on the web ?

A nice example with economic statistics for Amsterdam.

It works with cardboard, htc vive and oculus rift.

https://runhar.github.io/CityViewR/

.VR

What's next?

No one is seriously speaking about 2nd generation consumer VR headsets. The focus is about enhancing actual products (wireless, better controller, getting lighter, etc.) and cutting cost down toward mass market adoption.

Microsoft has been pushing assemblers (Asus, Lenovo, etc.) to release "cheapt" headsets (~400$) this winter.

Varjo big press release last week.

Only a prototype right now, with two super high density displays mounted on top of the Oculus Rift.

They want to achieve foveated rendering with "human eye resolution" at the center (~70 megapixels) and "regular" resolution on the side.

Phone solutions lack the power and the spatial tracking, but it's only a matter of time before it gets more than a gimmick.

.AR

What's next?

AR glasses technology is still a step behind VR in term of maturity. There is no real consumer grade device available on the market for the moment (it's mostly expensive dev kit/kickstarters).

New Hololens release in 2019?

Main contenders:

- Avegant Light Field

- Meta 2 headset

- Magic Leap

Avegant Light Field

Meta 2 prototype