• 検索結果がありません。

88 

(3)物体の全周を計測して得られた形状データから,物体の位相に関する事前知識をもっ ことなく任意の位相に適応して柔軟に全周型物体モデノレを生成する仕組みを実現した.

本手法では,等高面法と最適化手法を用いることにより,これまでモデリングの対象が 球面と同相な物体に限られていたのに対し,穴領域を有する物体や複数の物体からな

る三次元シーンなどに対するモデリングも可能になった.

謝 辞

本論文は,筆者が福岡県工業技術センター機械電子研究所職員として在職中および九州大 学大学院システム情報科学研究科知能システム学専攻博士後期課程在学中に行なった研究を まとめたものです.

本研究科において,知能システム学専攻長谷川勉教授には,本論文のまとめ方など終始親 身に多大な御指導,御助言および御鞭捷を賜わりました.ここに心からの謝意を表します.ま た,査紅彬助教授には,本研究を遂行するうえで,終始親身に多大な御指導,御助言および 御鞭掻を賜わりました.ここに心から感謝いたします.さらに,本論文をまとめるにあたり,

本研究科知能システム学専攻迫江博昭教授,同知能システム学専攻谷口倫一郎教授より有益 な御批評を頂きました.ここに心から感謝する次第です.

工学部情報工学科長田正教授(現九州システム情報技術研究所所長)には,終始親身に多 大な御指導,御助言および御鞭捷を頂きました.ここに心から感謝する次第です.

木室義彦講師(現九州、│システム情報技術研究所研究員),岡田伸鹿講師には,有益な御教示 と御助言を頂きました.お二人に心から感謝いたします.桐木利弘助手,江藤淳二技官には 有益な御助言と御協力を頂きました.お二人に心から感謝いたします.

九州大学大学院システム情報科学研究科長谷川研究室および査研究室の大学院生,学部生 および卒業生の皆様には熱心な御討論を頂きました.ここに深く感謝いたします.

松永純一氏(現飯塚研究開発機構研究開発部長)をはじめ,福岡県工業技術センター機械 電子研究所の先輩諸氏,同僚の皆様には,多くの御助言と常に暖かな励ましをいただきまし た.ここに深く感謝いたします.

最後に,博士課程に進学し,研究生活を送ることに対して深い理解を示し,応援していた だいた妻徳子と両親に心から感謝いたします.

1999年9月

参 考 文 献

[1]  L.  G. Ro berts. Machine perception of three‑dimensional solids

, 

optical and electro‑ optical information processing. Technical report

, 

MIT Press

, 

1965. 

[2]  A. Guzman. Decomposition of a visual scene into three‑dimensional bodies. In Proc.  of Fall Joint Comp. Con

  , . f

1968. 

[3]  M. Oshima and Y. Shirai.  Object recognition using three‑dimensional information.  1EEE Trans. PAλ11

, 

Vo1.5

, 

No

  , 4 .

pp.353361

1983. 

[4]  P.  Besl and R. Jain.  Invariant surface characteristics for  3D object recogtionin  range data. ComputerVisionGrαphicsαnd1mαge Processing

, 

Vo1.33

, 

pp.33‑80

, 

1986.  [5]  T. J.  Fan

, 

G. Medio

n爪1

tion. 1EEE Trαηs.  PAM1

, 

Vo.111

, 

No.11

, 

pp.11401157

1989. 

[6]  P.  J.  Besl and N. D. McKay. A method for registration of 3Dshapes. 1EEE Trans.  PAλ11

, 

Vol.14

, 

No.2

, 

pp.239256

1992. 

[7]  Y. Chen and G. Medioni. Object modelling by registration of multiple range images.  1mαgeαnd Vision Computing, Vol.10

, 

pp.145155

1992. 

[8]  H. Zha

, 

K. Morooka

, 

and T. Hasegawa. Active modeling of 3Dobjects:  planning on  the next best pose (NBP) for acquiring range images. In Proc. of 1nt.  Conf. on Recent  Advances in 3‑D Digital 1mαging and Modeling

, 

pp.6875

, 

1997. 

[9]  T. Poggio

, 

V. Torre

, 

and C. Koch. Computational vision and regularization theory.  Nature, Vo.1317

, 

pp.314‑319

, 

1985. 

[10]  A. N. Tikhonov and V. Y. Arsenin. Solutions of Ill‑Posed Problems. Winston and  Wiley Publishers. Washington D.C.

, 

1977. 

[11]  S.  Geman and D. Geman. Stochastic relaxation

, 

Gibbs distributions and the Bayesian 

restoration of images. 1EEE Trαns.  PAM1

, 

Vo1.6

, 

No.6

, 

pp.721741

1984. 

参 考 文 献 91 

[12]  J.  L. Marroquin. Deterministic Bayesian estimation of Markovian random fields with  applications  to  computational vision.  In Proc.  of Int.  Conf.  on Computer Vision

, 

1987. 

[13]  E.  B.  Gamble and T. Poggio. Visual integration and detection of discontinuities  :  The key role of intensity edges. A riificialIntell.  Lαb.

, 

Massαchusetts Inst.  Technol.

, 

AI Memo

, 

No.9701987.

[14]  T. Poggio

, 

E.  B.  Gamble

, 

and J.  J.  Little.  Parallel integra1

Science

, 

Vo.1242

, 

pp.436440

1988. 

[15]  S. Z. Li. Markov Random Field Modeling in Computer Vision. Berlin:Springer‑Verlag

, 

1995. 

[16]  S.‑Z. Li.  Towards 3D vision from range images:  An optimization framework and  parallel networks. ComputerVisionGraphics and Image Processing

, 

Vo1.24

, 

pp.52

96

, 

1983. 

[17]浦浜喜一,渡辺 真.画像の非線形尺度空間フィノレタリング:連続変形法と非等方拡散.

電子情報通信学会論文誌, Vol.J77D‑II,No.9, pp.1761‑1767, 1994. 

[18]  V. Granville

, 

M. Krivanek

, 

and J.  P.  Rasson. Simulated annealing:  A proof of con‑ vergence. IEEE 7αns.PAMI

, 

Vol.16

, 

pp.652656

1994. 

[19]  D. Geiger and F. Girosi. Parallel and deterministic algorithms from MRF's: Surface  reconstruction. IEEE Trαηs.  PAMI

, 

Vol.13, No.5, pp.401412,1991. 

[20]  G. Whitten. A framework for adaptive scale space tracking solutions to problems in  computational vision. In Proc. of Int.  Conf. on ComputerVision

pp.210‑220

, 

1990.  [21 ]横矢直和.多重スケールでの正則化によるステレオ画像からの不連続を保存した曲面再

構成.電子情報通信学会論文誌, Vo.lJ76D‑II,No.8, pp.1667‑1675, 1993. 

[22]  P. Craven and G. Wahba. Smoothing noisy data with spline functions ‑Estimating the  correct degree of smoothing by the method of generalized cross‑validation. Numerische  Mαthematik

, 

No.31

, 

pp.377403

, 

1979. 

[23]  B. Shahraray and D. J.  Anderson. Optimal estimation of contour properties by cross‑ validated regularization. IEEE Trans. PAMI

, 

Vo1.11, No.6, pp.600610,1989. 

参 考 文 献 92 

[24]  G. Wahba and J. Wendelberger. Some new mathematical methods for variational  objective analysis using splines and cross‑validation. Monthly Weather Rev.

, 

Vol.108

, 

pp.3657

1980. 

[25]  L.  D. Cohen and 

  . r

Cohen. Finite‑element methods for active contour models and  balloons for 2D and 3‑D images. 1EEE 1αns.PAM1

, 

Vol.15

, 

No.11

, 

pp.11311147

1993. 

[26]  D. Terzopoulos and M. Vasilescu. Sampling and reconstruction with adaptive meshes.  In Proc. of Computer Vision αnd Pattern Recognition

, 

pp.70‑75

, 

1991. 

[27]  H. Delingette

, 

M. Hebert

, 

and K. Ikeuchi. 8hape representation and image segmenta

tion using deformable surfaces. In Proc. of Compute1Visionαnd Pαttern Recognition

, 

pp

. 4

67472

1991. 

[28]  S.  Chen

, 

G. Stockman

, 

and K. Chang. 80 dynamic deformation for building of 3D models. 1EEE 1αηs.N N

, 

Vo1.7

, 

pp.374387

1996. 

[29]  K. Higuchi

, 

M. Hebert

, 

and K. Ikeuchi. Building 3Dmodels from unregistered range  images. In Proc. of 1nt.  Con

J .  

on Robotics and Automation

, 

pp.22482253

1994.  [30]  T. McInerney and D. Terzopoulos. A finete element model for 3D shape reconstruction 

and nonrigid motion tracking. In Proc. of 1nt.  Conf. 0η Compute1Vision

pp.518523

1993. 

[31]  A. J. Bulpitt and N. D. Efford. An efficient 3D deformable model with a self‑optimising  mesh.1mαge and Vision Computing

, 

Vo.l14

, 

pp.573580

1996. 

[32]  H. Hoppe

, 

T. DeRo T.Duchamp

, 

J. McDo

na 

tion. In P1oc.of S1GGRAPH 93

, 

Vo1.27

, 

pp.1926

1993. 

[33]  M. Kass

, 

A. Witkin

, 

and D. Terzopoulos. Snakes:  Active .contour models. 1nt. J.  01  Compute1Vision

Vo1.1

, 

No

  , 4 .

pp.321‑331

, 

1987. 

[34]  D. Terzopoulos

, 

A. Witkin

, 

and M. Kass. Constraints on deformable models: Recov‑ ering 3D shape and nonrigid motion. A 1tificial1ntelligence

, 

Vo.136

, 

No.1

, 

pp.91123

1988. 

[35]  R. Malladi

, 

J. A. Sethian

, 

and B. C. Vcm凶 .Shape modeling with front propagation:  A level set approach. 1EEE Trαηs.  PAM1

, 

Vo.l17

, 

No.2

, 

pp.158175

1995. 

参 考 文 献 93 

[36]  V. Caselles

, 

R. Kimmel

, 

and G. Sapiro.  Geodesic active contours. In Proc.  of 1nt.  Conf. on Compute1Vision

pp.69ι699

, 

1995. 

[37]  A. Blake and A. Zisserman. Visual Reconstruction

, 

Cambridge

, 

MA:MIT Press

, 

1987.  [38]  F. Spitzer. Markov Random Fields and Gibbs Ensembles. Amer.MIαth.Mon

, ・

Vo1.78

, 

pp.142‑154

, 

1971. 

[39]  D. Terzopo sand D. Metaxas. Dynamic 3D models with local and global deforma‑ tions:  Deformable superquadrics. 1EEE T:ns.PAM1

, 

Vol.13

, 

pp.703‑714

, 

1991.  [40]  D. Metaxas and D. Terzopoulos.  Shape and nongirid motion estimation through 

physics‑based synthesis. 1EEE 1αns.PAλ11, Vol.15, pp.580591,1993. 

[41]  D. Metax E.Koh

, 

and N. 1. Badler. Multi‑level shape representation using global  deformation and locally adaptive finite elements. Int. J. of Computer Vision

, 

Vo1.25

, 

No.1

, 

pp.4961

1997. 

[42] J.  Park

, 

D. Metaxas

, 

and A. Young. Deformable models with parameter functions:  Application to heart‑wall modeling. In Proc. 

0 1  

Computer Vision αnd Pattern Recog‑ nition

, 

pp.437‑442

, 

1994. 

[43]  D. Terzopoulos. Reg .rizationof inverse visual problems involving discontinuities.  1EEE Trαηs.  PAλ11, Vo1.8, No, .4 pp.413424,1986. 

[44]  G. Dhatt and G. Touzot. The Finite Element Method Displayed. Wiley

, 

New York

, 

1984. 

[45]  C. Rao. Adv α nced statistical methods iη

biωomηnetr TreωsseaαM7ア 川 .

[46]長尾 真監訳.ディジタノレ画像処理.近代科学社, 1972. 

[47]  P. W. M. Tsang

, 

P. C. Yuen

, 

and F. K. Lam. Recognition of occIuded objects. Pattern  Recognition

, 

Vo1.25

, 

pp.1107‑1117

, 

1992. 

[48]  D. Terzopoulos.  The computation of visible‑surface  representations. 1EEE Trans.  PAM1

, 

Vo.110, No.4,  pp.417438,1988. 

[49]  T. E. Boult and A. D. Gross. Recovery 0fse

Proc. of SP1E 1ntell.  Robots αnd Comp. Vision

, 

pp.358365

1987. 

参 考 文 献

94 

[50] Y. Sato

, 

K. Hattori

, 

and M. Ot叫<I.Real‑time handy rangefinder Cubicscope. In Proc.  of Int.  Conf. on Automation

, 

Robotics αnd ComputerVision

pp.11481152

1994.  [51]  S.  Kumar and D. Goldgof.  Model based part segmentation of range data: 'hyper‑

quadrics and dividing planes. In Proc.of WorkshopPhysics‑Bαsed M odeling in  Com‑

puter Vision

, 

pp.1723

1995. 

[52]  H. Zha, T. Hoshide, and T. Hasegawa, A recursive日ttingand splitting algorithm for 3

D object modeling using superquadrics. In Proc. of Int.  Conf. on Pattern Recognition

, 

pp.658662

1998. 

[53]  M. R. Korn and C. R. Dyer. 3D multiview object representations for model‑based  object recognition. Pαttern Recognition.

, 

Vo1.20, No.1, pp.91‑103, 1987. 

[54]坂根茂幸,石井 優,柿倉正義.ハンドアイ行動シミュレーク :'HEAVENシステムに基 づく視覚センサのオクルージョン回避.日本ロボット学会誌.Vo1.4, No.4,  pp.382392, 1986. 

[55J  D. J.  Williams and M. Shah. A fast algorithm for active contours. In Proc.  of Int.  Con

  f .

on Computer Vision

, 

pp.592595

1990. 

[56]  C. Koch

, 

J.  Marroquin

, 

and A. Yuille. Analog neuronal networks in early vision. In  Proc.of Natl. Acαd.  Sci.  USA

, 

pp.4263‑4267

, 

1986. 

[57]久長穣,山下 雅史,阿江忠.ホップフィールドニューラルネットワークによる実数 集合の分割.電子情報通信学会論文誌, Vol.J73‑D‑I ,I No.8, pp.12131219,1990.  [58]  S.  Osher  and  J.  A.  Sethian.  Fronts  propagating  with  curvature  dependent 

speed:Algorithms  based  on  Hamilton‑Jacobi  formulation.  J.  of  Computationαl  Physics

, 

Vol. 79

, 

pp.1249

1988. 

[59]  J.  A. Set1山n.Le SetMethods. Cambridge University Press

, 

1996. 

[60]  J.  A. Sethian and J.  Strain.  Crystal growth and dendritic solidification. ofCom‑

putationα1 Physics

, 

Vo1.98

, 

pp.231253

1992. 

[61 ]儀我 美ーほか.動く曲面を追いかけて.日本評論社, 1996. 

[62]  D. L. Chopp. Computing minimal surfaces via level set curvature flow. J. of Comp1

αttionα1 Physics

, 

Vo.1106

, 

pp.77‑91

, 

1993. 

参 考 文 献 95 

[63]  R. T. Whitaker. A level‑set approach to 3D reconstruction from range data. Int.よ of Compute1Vision

Vo1.29

, 

No.3

, 

pp.203231

1998. 

関連したドキュメント