{"id":9694,"date":"2017-02-25T11:48:13","date_gmt":"2017-02-25T10:48:13","guid":{"rendered":"http:\/\/www.keris-studio.fr\/blog\/?p=9694"},"modified":"2017-02-25T11:48:13","modified_gmt":"2017-02-25T10:48:13","slug":"3d-photo-ancient-photos-in-3d-models","status":"publish","type":"post","link":"https:\/\/www.keris-studio.fr\/blog\/?p=9694","title":{"rendered":"3D-Photo – ancient photos in 3D models"},"content":{"rendered":"

Placement de documents anciens dans une g\u00e9om\u00e9trie 3D<\/em><\/p>\n

\"\"<\/a><\/p>\n

It could be interesting to place ancient documents like drawings or photographs in a recent 3D model. It may open several uses like comparing before\/after or guess the photo has been taken from. It’s also possible to reconstruct a model from a collection of ancient photos. Other uses are yet to be determined.<\/strong><\/span><\/p>\n

Il peut \u00eatre int\u00e9ressant de travailler en 3D d\u2019apr\u00e8s des archives graphiques dont on ne maitrise pas toujours l\u2019origine. Cela peut avoir plusieurs usages, superposer des vues avant\/apr\u00e8s, replacer dans l\u2019espace des points de vue alors que l\u2019environnement a chang\u00e9, construire une photomod\u00e9lisation d\u2019une collection de photos ou encore projeter un \u00e9tat ant\u00e9rieur sur un \u00e9tat pr\u00e9sent. D\u2019autres usages sont encore \u00e0 d\u00e9terminer.<\/em><\/p>\n

If the work is based on a recent digital photo, things are much easier, but here we deal with old documents, old photographs or drawings. The most reliable technique, even if it remains tedious, is to find homologous points on the model and the photograph. It is a rather old process that preceded the automated operations of the recent <\/span><\/strong>software for photogrammetry. It can be done either in 3smax or Blender or other 3D graphic tools.<\/span><\/strong><\/p>\n

Si les documents \u00e0 utiliser sont au format num\u00e9rique, l\u2019op\u00e9ration peut \u00eatre assez simple et passer par un logiciel de photogramm\u00e9trie. Les g\u00e9om\u00e9tries anciennes et nouvelles peuvent \u00eatre assembl\u00e9es par points homologues. Si les documents sont partiellement transform\u00e9s ou tr\u00e8s anciens, il faut proc\u00e9der autrement.<\/em><\/p>\n

La technique la plus fiable, m\u00eame si elle reste fastidieuse, est celle qui consiste \u00e0 d\u00e9signer sur le mod\u00e8le puis sur la photo, les points homologues. C\u2019est un proc\u00e9d\u00e9 assez ancien qui a pr\u00e9c\u00e9d\u00e9 les op\u00e9rations automatis\u00e9es des logiciels r\u00e9cents de photomod\u00e9lisation. 3Dsmax poss\u00e8de les outils qui permettent de le faire, d\u2019autres outils comme Blender, ont des protocoles assez proches.<\/em><\/p>\n

Preparation<\/h2>\n

The first step consists on verifying the possibility of superposing the old photo and the 3D model by finding a similar view.<\/span><\/strong><\/p>\n

Dans un premier temps, un test est r\u00e9alis\u00e9 sur une vue ayant servi \u00e0 la photomod\u00e9lisation de l\u2019architecture. Il s\u2019agit de v\u00e9rifier dans un premier temps que l\u2019on peut retrouver la position parfaite, puisque les documents sont coh\u00e9rents.<\/em><\/p>\n

The 3D geometry is openned in 3DsMax :<\/span><\/strong><\/p>\n

Le mod\u00e8le 3D est charg\u00e9\u00a0:<\/em><\/p>\n

<\/p>\n

It’s no need to have a very high resolution to start.\u00a02000×1123 is here far enough.<\/span><\/strong><\/p>\n

La photo devant servir pour le calage est ramen\u00e9e dans un format plus l\u00e9ger 2000×1123, une r\u00e9solution plus grande n\u2019est pas n\u00e9cessaire.<\/em> <\/p>\n

Using\u00a0\u00ab\u00a0Camera Match\u00a0\u00bb in 3DsMax, match points are pinned on the 3D geometry.<\/span><\/strong><\/p>\n

Avec la fonction \u00ab\u00a0Camera Match\u00a0\u00bb les points de corr\u00e9lation sont positionn\u00e9s sur la g\u00e9om\u00e9trie\u00a0:<\/em><\/p>\n

<\/p>\n

<\/p>\n

Correspondances are set with\u00a0\u00ab\u00a0Camera Match\u00a0\u00bb<\/strong><\/span><\/p>\n

Puis les correspondances sont d\u00e9finies toujours avec la fonction \u00ab\u00a0Camera Match\u00a0\u00bb<\/em><\/p>\n

<\/p>\n

Once done, a camera has to be set. the 3D geometry is then automatically positioned to match the photo.<\/strong><\/span><\/p>\n

Une fois fait, il faut cr\u00e9er une cam\u00e9ra et le mod\u00e8le est positionn\u00e9 dans la perspective de la photo.<\/em><\/p>\n

<\/p>\n

The Matching is perfect, it’s now possible to find the original location of the camera when the photo has been taken.<\/strong><\/span><\/p>\n

La superposition est parfaite, ce qui permet de retrouver la position de la cam\u00e9ra au moment de la prise de vue.<\/em><\/p>\n

En semi transparence\u00a0: <\/p>\n

<\/p>\n

This first test was an easy one. The photo was recent and the matching process was easy. Let’s try now with an old document taken from a PDF file. EXIF information\u00a0is lost or have never existed for analog photographs.<\/strong><\/span><\/p>\n

In this second test, the view is taken from the PDF file\u00a0\u00ab\u00a0Autour de la Table<\/a>\u00a0\u00bb.\u00a0<\/em><\/strong><\/span><\/p>\n

Dans cette seco<\/strong><\/span>nde v\u00e9rification, la photo est extraite du rapport au format PDF \u00ab\u00a0Autour de la Table<\/a>\u00a0\u00bb. Les images sont vraisemblablement recadr\u00e9es et ont perdu toutes informations concernant la prise de vue.<\/em><\/p>\n

<\/p>\n

The size of the photo is\u00a0327×494 and has been put in background. The rendering size has been set with the same proportions.<\/strong><\/span><\/p>\n

La photo au format 327×494 est plac\u00e9e en fond d\u2019\u00e9cran. Et les param\u00e8tres de rendu sont r\u00e9gl\u00e9s dans les m\u00eames proportions.<\/em><\/p>\n

<\/p>\n

The background matches the output to avoid deformations.<\/span><\/strong><\/p>\n

L\u2019image en fond d\u2019\u00e9cran est param\u00e9tr\u00e9e de telle fa\u00e7on que son affichage correspondra aux proportions du rendu. Cela afin d\u2019\u00e9viter toute d\u00e9formation.<\/em> <\/p>\n

It’s now time to choose the matching points, precision is needed.<\/span><\/strong><\/p>\n

On d\u00e9termine la position des points, il faut \u00eatre pr\u00e9cis car il n\u2019y aura plus d\u2019indications ensuite.<\/em><\/p>\n

<\/p>\n

One done, the pins have to be positioned on the 3D model. There, one can notice that one stone is missing under the\u00a0lintel, it may have fell, knowing that the cairn is a recent reconstruction.<\/strong><\/span><\/p>\n

Une fois le choix fait, il faut placer les rep\u00e8res sur le mod\u00e8le g\u00e9om\u00e9trique. (On remarque qu\u2019il manque une pierre sous le linteau, sur la photo \u00e0 gauche.)<\/em><\/p>\n

<\/p>\n

In the viewing window, the texture is in low resolution. It makes the positioning\u00a0<\/strong>more difucult. Then, once again, it’s needed the match the pinpoints.<\/span><\/strong><\/p>\n

La texture \u00e9tant montr\u00e9e en basse d\u00e9finition, le jeu reste p\u00e9rilleux. Il faut ensuite positionner sur la photo plac\u00e9e en fond d\u2019\u00e9cran, la correspondance de chaque rep\u00e8re.<\/em><\/p>\n

<\/p>\n

Then, a camera has to be created. If it matches, it is won. If not, it may be necessary to refine. It’s important to find angles.<\/strong><\/span><\/p>\n

Une fois l\u2019op\u00e9ration termin\u00e9e, il faut cr\u00e9er une cam\u00e9ra. Si le mod\u00e8le se positionne bien, c\u2019est gagn\u00e9, sinon, il faut repenser la position des points, il est important qu\u2019il y ait des angles.<\/em><\/p>\n

Checking the results with transparency:<\/strong><\/span><\/p>\n

La vue est recal\u00e9e, v\u00e9rification en transparence\u00a0:<\/em><\/p>\n

— <\/p>\n

It’s now possible to guess where the original photograph has been taken from and nearly recreate the EXIF information.<\/span><\/strong><\/p>\n

On peut d\u00e8s lors d\u00e9terminer la position du photographe\u00a0et par extention le r\u00e9glage de son appareil au moment de la prise de vue. Mais dans cet exemple, le document est tellement modifi\u00e9, que les erreurs sont probables :<\/em><\/p>\n

<\/p>\n

By multiplying the method on a survey campaign, it would be possible to detect anomalies or details that had slipped from initial analysis.<\/strong><\/span><\/p>\n

En multipliant les exemples sur une campagne de fouille, il serait possible de replacer une bonne partie des vues et d\u2019\u00e9ventuellement d\u00e9tecter des anomalies ou d\u00e9tails qui avaient \u00e9chapp\u00e9 \u00e0 une premi\u00e8re analyse.<\/em><\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

Placement de documents anciens dans une g\u00e9om\u00e9trie 3D It could be interesting to place ancient documents like drawings or photographs in a recent 3D model. It may open several uses like comparing before\/after or guess the photo has been taken from. It’s also possible to reconstruct a model from a collection of ancient photos. Other … Continuer la lecture de 3D-Photo – ancient photos in 3D models<\/span> →<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6,7,117,8,147],"tags":[301,271,302,308,244,325,276,311,272],"class_list":["post-9694","post","type-post","status-publish","format-standard","hentry","category-3d","category-3dsmax","category-archeologie","category-methodologie","category-photomodelisation","tag-3d","tag-3ds-max","tag-3dsmax","tag-archeologie","tag-archeology","tag-bretagne","tag-photogrammetry","tag-photomodelisation","tag-table-des-marchands"],"_links":{"self":[{"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/posts\/9694","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=9694"}],"version-history":[{"count":2,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/posts\/9694\/revisions"}],"predecessor-version":[{"id":9860,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=\/wp\/v2\/posts\/9694\/revisions\/9860"}],"wp:attachment":[{"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=9694"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=9694"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.keris-studio.fr\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=9694"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}