Title: Algorithm for Fusion of 3D Scene by Subgraph Isomorphism with Procrustes Analysis
1Algorithm for Fusion of 3D Sceneby Subgraph
Isomorphism with Procrustes Analysis
- Krzysztof Skabek, Przemyslaw Kowalski
- Instytut Informatyki Teoretycznej i Stosowanej
PAN, - ul. Baltycka 5, 44-100 Gliwice
- e-mail krzysiek_at_iitis.gliwice.pl
2Contents
- 1. Active vision
- 2. Stages of 3D fusion
- 3. Graph representation and algorithms
- 5. Matching structure graphs
- 4. Algorithm for 3D Fusion
- 5. Experiments
3Active Vision Platform
4Purpose
-
- Obtaining a complete 3D representation of the
scene and relations between components of the
scene basing on a set of 3D frames from multiple
viewpoints. - Assumptions
- No a-priori information about objects of the
scene. - Unknown location of the viewpoint.
- We focus on polyhedral objects.
5Architecture of Active Vision
Sensing
Planning
Mision planning
3D model
Navigator
Location x,y,z,?
Pilot
Model integration
Image preprocessing
Act
Controller
camera
camera
Engines
63D Fusion of Multipoint Views
3D Fusion integration process of objects in 3D
scene on the basis of visual information from
several viewpoints.
7Stages of 3D Fusion
Vision device
Current view
3D model
Viewpoint loc.
Prediction
Corection
Knowledge of the scene
Navigation to a new viewpoint
8Preprocessing of Visual Information
- Improvement of image quality, noise reduction
- Image segmentation, extraction of lines,
segments,vertices Susan, Hough Methods - Stereo matching, depth map active contours,
hardware support (ranging lasers, depth sensors) - Algorithms prepared for Khoros platform
9Viewpoint parameters
- T vector of translation (31),
- R rotation matrix (33, orthogonal)
- s scale (scalar)
Relation between coordinates of point P Pw
global coordinates, Pk coordinate system of
the camera Pk R(Pw T)s
10Graph representation of 3D scene
4
7
Contour graph
4
6
1
3
1
3
5
2
2
Face graph
11Graph Isomorphism
Graph Isomorphism
?
3
3
2
2
2
2
3
3
12Detection of Graph Isomorphisms
- Permutation method
- Clique detection method
- Ullman method
- A method (error correction)
- Decision tree method
Algorthm with analysis of 3D structure
deformation (decision tree, consistency checking,
branch pruning, geometric similarity)
13Similarity of Shape Procrustes analysis
Rotation (R)
Translation (T)
Scale (S)
--- - object A --- - object B --- - A to B
matching D2(A, B) B SRA T 2
14Implementation of 3D Fusion(matching contour
graphs)
- Stage I
- Generation of groups of vertices (quadruplets)
fulfilling conditions - Procrustes distance lt ?
- Preserving edge topology
V
- Stage II (for eqch group of vertices from stage
I) - Calculation of local transformation (TL RL SL)
- Matching the remaining vertices
- Local Procrustes distance lt ?
- Preserving edge topology
- Calculation of exact transformation (T R S)
15Implementation of 3D Fusion(model updating)
1
2
B
F
3
A
D
5
4
C
6
E
GIi
GMi-1
16Implementation of 3D Fusion(hypothesis of the
scene objects)
- Introducing unconfirmed elements.
- Hypothesis of scene objects
- Connecting edges
- Closing faces
- Connecting partial faces
- Ground plane detection
- Completing vertical faces
17Conditions of experiments
- Total transformation error consists of
- rotation, translation, scale
- Tolerance of rotation (R?? matrix of rotation
error) - R R??R??
- Estimation of rotation error
- ? ½ 1 cos(??) ½
- Assumed value of rotation error
- ? 0.1 for ?? ? 16
18Experiments
19Thank you for your attention
20Graph representation of 3D scene II
- a set of vertices in the scene
- a set of edges between vertices
- coordinates (x,y,z) of vertices
- a set of faces in the scene
- a set of connections between faces
- parameters of faces
- parameters of connections between faces
21Implementation of 3D Fusion
- Input data
- GIi contour graphs for views (i1..n)
- Ti Ri Si estimated transformation (from
navigation unit) - ?i transformation tolerance (for
navigation unit) - ?i observation tolerance (for optical
unit) - Output data
- GMn Contour graphs of model
- Ti Ri Si computed transformation
- First step of fusion
- GM1 ? GI1
22Experiment I