Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Photogrammetrie: Le process
En 4 étapes:
- Capture
- Points homologues
- Positions relatives / absolues
- Reconstitution 3D
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
1. Capture
capturer des données
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Matériel de base : photographies
La photographie est une technique permettant de fixer une image sur une surface sensible à l’aide de la lumière et des objets qui la reflètent.
→ Large gamme de données de base
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Capteur pour photo aérienne IMAO B66
(faibles parallaxes)
- 18 x 50 Mpixels → 770Mpix
- 1 pixel = 6µm
- à 1000m, 1pixel = 3cm au sol
- 4Gb / file
- 1.8 frame/sec
→ Gagner du temps de vol (cher)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3372992/b66.jpg)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3372994/b66_1.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Mais pour nous:
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3372996/Nokia-6230i.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
2. Détection de zones homologues
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3372998/xkcd.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Plusieurs algorithme:
- 1999, SIFT: Scale Invariant Feature Transform
- 2008, SURF: Speeded Up Robust Features
- 2016, autres? (RealityCapture…)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
SIFT
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373000/IMG_20161208_083324.jpg)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373001/IMG_20161208_083319.jpg)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373002/IMG_20161208_083314.jpg)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373003/IMG_20161208_083307.jpg)
Points homologues sur différentes images
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373010/PANO.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
SIFT = Scale Invariant Feature Transform
- invariant à l’échelle
- invariant à la rotation
- invariant à l’exposition
→ Basé sur des caractéristiques locales
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
SIFT: algorithme robuste
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373013/robust.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
4 étapes pour SIFT:
- Détecter des zones caractéristiques
- Détecter leur orientation dominante
- Calculer leur descripteur
- Retrouver des zones similaires sur d’autres images
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
2.1. Zones caractéristiques
Descripteurs de Harris (1981)
= analyse locale de l’image à l’ordre 2
Opérateur de dérivation:
- DoG (Difference of Gaussians)
- LoG (Laplacian of Gaussian)
- DoH (Difference of Hessians)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Rappel Gaussienne
par exemple:
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373023/gaussienne3D.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Convolution = “remplacer” chaque pixel par son voisinage x matrice gaussienne
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373024/convolution.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Exemple de noyau de convolution gaussien
On y retrouve notre courbe 3D précédente
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373027/convolution2.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373028/DOG_Mexican.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373029/DOG_Mexican2.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373030/DOG_Mexican3.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373034/BaseGauss.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Avantage (par rapport à un noyau de convolution "moyenne"):
élimine les détails de haute fréquence
(= le bruit)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Descripteurs de Lowe (1990)
→passage à l’octave
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373045/octave.png)
→ parametre de précision
= jusqu'à quel octave descend-on?
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
2.2. Orientation
- Histogramme des orientations des gradients des pixels voisins dans Lσ
- Orientation = orientation correspondant au plus haut pic de l’histogramme
- Si plusieurs pics importants, on dédouble le point
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373056/orientationMarker.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
2.3. Descripteur
Caractériser le comportement de l’image au voisinage du point
- orientation des gradients de pixels
- histogramme de ces orientations au voisinage
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373057/histograme.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
4x4x8 = 128
descripteur du point: vecteur de dimension 128
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373062/descriptor.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373067/sift_basic_3.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
2.4. Matcher chez les autres
- Descripteur identique
- Proches voisins (économique)
- Alignements:
- transformées de Haught
- moindre carrés
- etc…
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Filtrage
On retire:
- les zone à faible contraste
- les arrêtes
- les aberrations
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373070/Sift_keypoints_filtering1.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373071/Sift_keypoints_filtering2.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373074/Sift_keypoints_filtering3.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Principe de filtrage des abbérations
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373078/filter.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Points détécté par SIFT
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373082/SIFT_ciel.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Points concordants
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373084/SIFT_ciel_match.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3. Positions absolue des prises de vue
- Concerne un couple stéréoscopique: 2 images avec recouvrement
- À partir des points communs
- ensemble de transformations matricielles
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3 étapes:
- Orientation relative
- Orientation relative + échelle
- Orientation absolue
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3.0. Corrections des distortions
- Via les caractéristiques de l’objectif
- Corrélation dense…
Souvent une combinaison des 2.
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3.1. Orientation relative
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373087/OrientationRelativeAbsolueMoisset2Orient.gif)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3.2. Orientation relative + échelle
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373089/OrientationRelativeAbsolueMoisset2Scaled.gif)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
- Cette semaine nous nous arrêterons là.
- Pas de points connus,
- Uniquement des points “connus” sur les autres couples stéréophotgrammetrique.
→ Échelle relative
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
3.3. Orientation absolue
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373091/OrientationRelativeAbsolueMoisset2GeoRef.gif)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
4. Reconstitution 3D
Prenons le cas de ce cube
Je prends des photos
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373098/cubeBlanc.png)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373092/cubeBlanc.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Différence des Gaussiennes
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373102/cubeBlancGauss.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Points qui matchent
Orientation des photos
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373107/cubeBlancGaussClean.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Comment obtenir tous les points en 3D ??
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3373111/cubeBlancToFind.jpg)
Matthieu QUANTIN 2016
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330154/Cc-by-nc-nd_icon.png)
TP photogrammetrie - Phycité
![](https://s3.amazonaws.com/media-p.slid.es/uploads/626751/images/3330192/ECN.png)
Photogrammetrie_Process
By matthieuquantin
Photogrammetrie_Process
- 906