Archives de catégorie : Méthodologie

Photogrammetry – reconstruction – 3Dpdf export – Part1

Set-up

The photogrammetry has been done with a very small setup. The Sony DSC QX10 has been used for its very small size which made the process easier.

Sony DSC-QX10 https://dyw7ncnq1en5l.cloudfront.net/optim/produits/16/17125/QX10_with_XperiaZ_2-1200-5226f5d11e322__450_400.jpg

This camera is connected to a smartphone via Wi-Fi, it is then possible to shoot from a distance.

The camera has been fixed on a small tripod for the lateral views and on a Smatree SmaPole Q3S Extensible for the upper views.

Résultat d’images pour Smatree SmaPole

Two rulers for archaeology have been at a 90° angle to indicate the horizontal plane and the scale. This is one of the most important part of the preparation.

The light has to be diffused, natural if possible. It’s important to have a very homogeneous light inside the model.

Capturing process

It’s important to start with a view that shows the model in total with the scale. There, the upper view has been chosen.

Then process as usual with a shape of globe around the model

It’s important to be very low to get information inside the model and at the base of it. If possible the Depth of field (DOF) has to be set at 16 or more, depending to the light. The more it is the more the image is focused everywhere. Important never zoom, move the camera instead. That’s why this small camera is useful as it can be very close even inside the model.

Put the model upside down and process again.

The photos can be sent to RECAP in order to get a complete model with online power resources.

Recap is an Autodesk service that changes a lot. Right now, there is Recap and Recap Photo. https://www.autodesk.com/products/recap/overview

The benefit of recap is to have an online process and a complete model with texture within a short time. Les options are let for tweaking or adapting the quality to the final needs. Anyway, it’s often a good choice mainly with less powerful computers.

Problem is, the reconstruction of the whole model has to be done manually.

Photoscan

In Photoscan, the process is roughly straightforward.

Step 1 aligning

Then

Getting a dense cloud

All parameters are set to High

It is possible at the stage to scale and fix the horizontal plane.

Scaling

For that click Right Mouse Button and Create a first Marker

On the photographs it is necessary to check if the marker is at a correct position, if not move it.

3 markers are needed in order to have a correct scales model and a correct horizontal plane;

Once done, assign the correct values to the points.

In reference

Then click update !!!!

Meshing

Next step is Meshing

Then texturing

Everything is perfect, even inside the model

The same process has to be repeated for the model upside down

Processing the other side

The best is to create a new chunk, rename and repeat the process

Assembling

The 2 chunks are duplicated

Then the ground and ruler are erased.

The two models are perfectly scaled but now they have to be aligned.

Alignment

Erase the previous markers and add some at specific location that can be found similar between the top view and bottom view.

The more Markers are positioned, the more accurate the model is.

Then in the Workflow tab, choose Align Chunks

When done chunks could be merged

Before doing that, it’s better to clean up a little the points that are a little bit out of the model.

Merging

It’s better to merge just Dense Clouds

Then Build mesh again

And textures

Depending the texturing mode, textures could be slightly different. Here, the spherical mode has been chosen.

Exporting

Export model to OBJ file

End of part 1

Figurations de l’intangible

Les invariants

Il y a-t-il, en arrivant au terme de cette exploration de figurations remarquables, où les phénomènes d’ambiance sont représentés, des invariants, des images prototypiques pour le vent, le soleil et la lumière, la température et le son ? Si ces prototypes existent, ils pourront être utilisés dans la construction des topiques d’ambiance (voir le chapitre concernant ce thème) et plus généralement dans les phases de référencement du projet, en accompagnement d’autres éléments graphiques ou textuels.

Les invariants sont en quelque sorte les atomes d’ambiances, les ambiantités comme cela a déjà été nommé plus haut. C’est ce qui reste lorsque l’on retire à l’image son principe de composition et de cadrage et la co-présence des séries. On retrouve donc, le mode d’expression graphique propre et les couleurs. Continuer la lecture de Figurations de l’intangible

Le tracé de Crucuno

Crucuno, photo Laurent Lescop
Crucuno, photo Laurent Lescop

Le site de Crucuno a-t-il été conçu comme l’a restauré Gaillard, ou ce dernier a-t-il régulé le tracé ? Dans le Archaeologia Cambrensis[1], il est dit que « The Cromlech of Crucuno now, after being « très bien restauré et dégagé», consists of twenty-one menhirs arranged in a rectangle, 136 ft. long by 96 ft. broad. » (page 72). Très bien, signifie-t-il fidèlement ? Outre les photos publiées par Gaillard, il « existe un plan établi par Muray Vicars en 1834, un relevé coté en 1867, visible au Musée de Guernesey, par Dryden et Lukis, puis un autre relevé, assez différent que l’on trouve daté de 1870 et parut avec une petite description dans le Wiltshire Archeological and Natural History[2] n°13, datant de 1872, comme déjà mentionné. Le plan suivant date de 1897, c’est celui de Gaillard que l’on trouve dans son Astronomie Préhistorique. Il faut attendre 1973 pour voir un nouveau relevé, c’est cela de Thom établi en 1973. En 2014, j’ai fait un relevé 3D qui sera présenté ensuite. Continuer la lecture de Le tracé de Crucuno