International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
of the coal pile is about 20 meters and the height is about 4
meters, and the shape is very irregular, so we took 83 photos
which just form a circle. All photos are taken by hand-held
camera and then input into computer and processed by our
software and 3D results are produced as shown in Fig. 3. Except
taking photos, all other processes are automatically executed by
computer. Results show reliability and robustness of our image
matching algorithm and convenience and efficiency and
automation level of our 3D reconstruction algorithm.
Fig. 3 Images and 3D Models of Reconstruction
S CONCLUSIONS
This paper proposed a new image matching method which uses
short baseline to maintain reliability of matching and uses multi
images to maintain intersection accuracy. Based this matching
technique, a 3D reconstruction system has been developed
which has characteristics of being easy using, fast and automatic
and can be widely used in 3D reconstruction of objects. Here a
metric camera is used. When using ordinary camera, camera
calibration process is essential or self-calibration algorithms
should be integrated into current calculation process. All these
arc our major research subjects.
Reference:
Li, D., Zheng, Z. Analytical photogrammetry, Publishing house
of survey and mapping, 1992.
Zhang, Z., Zhang, J. Digital photogrammetry. Publishing house
of survey and mapping, 1996.
Pollefeys, M.. Self-calibration and Metric 3D Reconstruction
from Uncalibrated Image Sequences. Doctor’s dissertation.
Katholieke University, Belgium,1999.
DURAN, Z., TOZ, Goniil. Obtaining 3D Information of
Historical Monuments by Means of Photogrammetry.
MONTI, Carlo. Laser Scanner Application on Complex Shapes
of Architecture. Profiles Extraction Processing and 3D
Modeling. [International Archives of the Photogrammetry,
Remote Sensing and Spatial Information
Sciences Vol. XXXIV-5/W10.
Delorme, S., Petit, Y., Guise, J.A.de. Three-Dimensional
Modeling and Rendering of the Human Skeletal Trunk from 2D
Radiographic Images.
Zhang, Y. Three Dimensional Reconstruction and Visual
Inspection of Industrial Sheetmetal Parts with Image Sequences.
Ph.D. dissertation, Wuhan University, 2002.
D'Apuzzo, N. Human face Modeling from Multi Images.
PowerPoint of Oral Speech in 3rd International Image Sensing
Seminar on New Development in Digital Photogrammetry,
September 2001, Gifu, Japan.
Sengupta, K. and Chi, C. Scanning Face Models with Desktop
Cameras. IEEE Transactions on Industrial Electronics, Vol
48,No 5,October 2001: 904-912.
Clarkson, J. Matthew, Rueckert, Daniel, Derek, L.G. Using
Photo-Consistency to Register 2D Optical Images of the Human
Face to a 3D Surface Model.IEEE Transaction on Pattern
Analysis and Matching Intelligence. Vol.23.No.11.November
2001:1266~1280.
Remondino, Fabio. Human Body Reconstruction from Image
Sequences. Pattern Recognition (DAGM 2002), Lecture Notes
in Computer Science 2449, Springer 2002: 50-57.
Zurich(Switzerland), September 2002.
Tashiro, K. Processing and Visualization of 3D Point Clouds
Data Capture by Laser Scanner Systems. PowerPoint of Oral
Speech in 3rd International Image Sensing Seminar on New
Development in Digital Photogrammetry, September
2001,Gifu,Japan.
Ulas Yilmaz, Adem Miilayim, Volkan Atalay. Reconstruction
of Three Dimensional Models from Real Images. Proceedings
of the First International Symposium on 3D Data Processing
Visualization Transmission (3DPVT'02). IEEE, 2002.
8 Dej
KEY V
ABSTEI
A high
method
the cen
is a sim
sensor :
in term
exposut
In. indt
precisic
configu
is stron:
is limite
after su
reduced
Han
paramet
adjustm
off-the-
frequen
cheap,
indisper
A conv
adjustm
3D con
taken in
be a go
with thi
well cal
sensor i
is captu
not onl
measure
This pay
adjustm
basic id
measure
field at
centre.
function
