3D multispectral acquisition

minorbigarmΑσφάλεια

30 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

122 εμφανίσεις

3D
MULTISPECTRAL

ACQUISITION


Steven Le Moan, Alamin Mansouri

Université de Bourgogne, Auxerre, FR


Workshop on Multispectral imaging and Biometrics

Gjøvik University College


November 2010

O
UTLINE


Context & goals


3D multispectral scanner


Depth imaging


Space carving


Perspectives

C
ONTEXT

&
GOALS


How to describe a real
-
life object ?


Shape, color, touch, smell, sound,..



How does it look like ?


Shape + Color >> 3D multispectral

C
ONTEXT

&
GOALS


What is a 3D multispectral dataset ?


Way too many dimensions !


One “pixel” is defined by:

-
3 spatial coordinates (X,Y,Z)

-
N spectral values (B
1
, B
2
, ..., B
N
)


>>
Multimodal (N+3) dimensional space

3D
MULTISPECTRAL

SCANNER


System overview










Mansouri A.,
Etude, conception et réalisation d’un système multispectral de vision pour des applications de
proximité et développement d’algorithmes de reconstruction de la reflectance,
PhD Thesis, 2005


3D
MULTISPECTRAL

SCANNER


Calibration










Caltech Calibration Toolbox for Matlab

3D
MULTISPECTRAL

SCANNER


Acquisition

3D
MULTISPECTRAL

SCANNER


Results











3D
MULTISPECTRAL

SCANNER


Results











3D
MULTISPECTRAL

SCANNER


Limitations (to be improved)


-

Lack of convenience

-

Heavy calibration

-

What about pearlescence ?


D
EPTH

IMAGING


Principle








The depth is printed in the blanket.


Page D.,
Part decomposition of 3D surfaces
, PhD thesis, 2003


D
EPTH

IMAGING


LIDAR
(LIght Detection And Ranging)

-
Remote sensing applications

-
Short wavelengths (UV


nIR)







Time
-
of
-
flight cameras

-

Close
-
range applications (up to 60m)

-
“One shot” acquisition











D
EPTH

IMAGING


Examples












Roy M.,
Comparaison et analyse multirésolution de maillages irréguliers avec attributs d’apparence,
PhD Thesis, 2004

D
EPTH

IMAGING


Examples












Image provided by Norsk Elektro Optikk

S
PACE

CARVING


Reconstruction from a set of images

P
ERSPECTIVES


Convenience (“point & shoot”)



Fast calibration



Unified framework ?

T
HANK

YOU

FOR

YOUR

ATTENTION