the roughness
mong the ob-
1 region match
th the highest
1umber of ver-
as the largest
finally selected
e pose of the
ct model is vi-
ing phase, the
ained by com-
vertices in the
nding vertices
works to solve
ence problems
ed scheme can
1 process. In a
multiple-view
of 2-D projec-
By calculating
ie input image
abase ,a set of
ng score is se-
| as a coarse
cted from the
eld net for es-
etween the in-
is phase is the
that has the
S with the in-
bject model is
inate frame as
se of the un-
nethods , there
d nets for im-
regarded as a
veen two rela-
NP-complete
nplexity is ex-
speed up the
look-ahead or
een proposed.
ort to manage
algorithm. A
its massively
g problems in
quantitatively
final states of
etwork struc-
is easy to im-
REFERENCES
[1]P. J. Besl and R. C. Jain, “Three-dimensional
object recognition”, ACM Comput. Surveys,
vol. 17, no. 1, pp. 75-145, Mar. 1985.
[2]M. Oshima and Y. Shirai, “Object recognition us-
ing three-dimensional information ”, IEEE
Trans. Pattern Anal. Machine Intell. , vol. PA-
MI-5, no. 4, pp. 353-361, July 1983.
[3]R.T. Chin and C.R. Dyer, “Model-based recog-
nition in robot vision”, ACM Comput. Surveys,
vol. 18, no. 1 pp. 67-108, Mar. 1986.
[4]I. Chakravarty and H. Freeman, "Characteristic
views as a basis for three-dimensional object
recognition”, in Proc. Soc. Photo-Opt. Instr.
Eng , vol. 336, pp. 37-45, 1982.
[5]H. Freeman and I. Chakravarty, “The use of
characteristic views in the recognition of three-di-
mensional objects”, E. Gelsema and L. Kanal,
Eds. , Pattern Recognition in Practice. Amster-
dam: The Netherlands, North-Holland, pp.
277-288,1980.
[6]L. Stark, D. Eggert, and K. Bowyer, “Aspect
graphs and nonlinear optimization in 3-D object
recognition”, IEEE 2nd Int. Conf Computer V i-
sion, Tampa, FL. Dec. 5-8, 1988, pp. 501-
507.
[7]Y. Zhang, “An automated procedure to construct
minimum aspect graphs for 3-D object recogni-
tion", Journal of Pattern Recognition, Vol. 10,
pp. 132-141,1994.
[8]J. J. Hopfield and D. W. Tank, "Neural computa-
tion of decisions in optimization problems”, Bi
olog. Cybernet. , vol. 52, pp. 141-152, 1985.
[9]—, “Computing with neural circuits; A model",
Science, vol. 233, pp. 625-633, 1986.
[10] W. Li and M. Nasrabadi, "Object recognition
based on graph matching implemented by a hop-
field-style neural network”, Int. J. Conf. Neu-
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
ral Networks, II. Washington, DC, 1989.
[11]B. Parvin and G. Medioni, “A constraint satis-
faction network for matching 3D objects”, Int.
J. Conf. Neural Networks, II. Washington,
DC, 1989.
[12]E. Mjolsness, G. Gindi, and P. Anandan, "Op-
timization in model matching and perceptual or-
ganization”, Neural Computation, vol. 1, pp.
218-229,1989.
[13]R. Mohan, "Application of neural constraint
satisfaction networks to vision", Int. J. Conf.
Neural Networks,II. Washington, DC ,19689.
[14] P. J. Bes! and R. C. Jain, "Segmentation
through variable-order surface fitting”, IEEE
Trans. Pattern Anal. Machine Intell. , vol. PA-
MI-10, no. 2, pp. 167-192, Mar, 1988.
[15]S. Y. Chen and M. Y. Chern, "Flexible corner
detection based on a single-parameter control",
Proc. of SPIE, vol. 848, pp. 108-114,1987.
[16]S. A. Underwood and C. L. Coates, Jr. , “Visu-
al learning from multiple views", IEEE T rans.
Com put. , vol C-24, no. 6, pp. 651-661, 1975.
[17]G. W. Davis, "Sensitivity analysis in ncural net
solutions”, IEEE Trans. Syst., Man, Cyber-
net. , vol. 19. no. 5, pp. 1078-1082, Sept. /
Oct. 1989.
[18]L. G. Shapiro and R. M. Haralick , “Structural
description and inexact matching”, IEEE Trans.
Pattern Anal. Machine Intell. , vol. PAMI-3,
pp. 504-515, 1981.
[19]L. G. Shapiro, J. D. Moriarty, and R. M. Haral-
ick , “Matching three-dimensional objects using a
relational paradigm”, Pattern Recognwion, vol.
17, no. 4, pp. 385-405, 1984.
[20]R. M. Haralick and J. S. Kartus, "Arrange-
ments ;,homomorphisms ,and discrete relaxation" ,
IEEE Trans Syst. , Man, Cybern. , vol. SMC-
8, pp. 600-612, Aug. 1978
1017
