Bisser Raytchev. 739-8527 1-4-1 630-0192 8916 5 E-mail: {tamaki,bisser,kin}@hiroshima-u.ac.jp, amano@is.naist.jp. R n R



Σχετικά έγγραφα
INTEGRATION OF THE NORMAL DISTRIBUTION CURVE

Homework for 1/27 Due 2/5

Rapid Acquisitio n of Doppler Shift in Satellite Co mmunicatio ns

p n r

LAD Estimation for Time Series Models With Finite and Infinite Variance

K. Hausdorff K K O X = SDA. symbolic data analysis SDA SDA. Vol. 16 No. 3 Mar JOURNAL OF MANAGEMENT SCIENCES IN CHINA

Research on Real-Time Collision Detection Based on Hybrid Hierarchical Bounding Volume

1. For each of the following power series, find the interval of convergence and the radius of convergence:

J. of Math. (PRC) Shannon-McMillan, , McMillan [2] Breiman [3] , Algoet Cover [10] AEP. P (X n m = x n m) = p m,n (x n m) > 0, x i X, 0 m i n. (1.

On Generating Relations of Some Triple. Hypergeometric Functions

Pro duction Technology and Technical Efficiency in ( k, y) Sp ace

CDMA. Performance Analysis of Chaotic Spread Spectrum CDMA Systems. LI Xiao - chao, GUO Dong - hui, ZENG Quan, WU Bo - xi RESEARCH & DEVELOPMENT

An A lgor ithm of M ea sur ing Var ia tion D egree for Fea ture M odel

Detection and Recognition of Traffic Signal Using Machine Learning

SUPERPOSITION, MEASUREMENT, NORMALIZATION, EXPECTATION VALUES. Reading: QM course packet Ch 5 up to 5.6

Ψηφιακή Επεξεργασία Εικόνας

þÿ Ç»¹º ³µÃ ± : Ãż²» Ä Â

Outline. Detection Theory. Background. Background (Cont.)

On Inclusion Relation of Absolute Summability

ΕΛΕΓΧΟΣ ΤΩΝ ΠΑΡΑΜΟΡΦΩΣΕΩΝ ΧΑΛΥΒ ΙΝΩΝ ΦΟΡΕΩΝ ΜΕΓΑΛΟΥ ΑΝΟΙΓΜΑΤΟΣ ΤΥΠΟΥ MBSN ΜΕ ΤΗ ΧΡΗΣΗ ΚΑΛΩ ΙΩΝ: ΠΡΟΤΑΣΗ ΕΦΑΡΜΟΓΗΣ ΣΕ ΑΝΟΙΚΤΟ ΣΤΕΓΑΣΤΡΟ

Homework 3 Solutions

PACS: Pq, Tp

A Vision Based Method for Aircraft Approach Angle Estimation

Congruence Classes of Invertible Matrices of Order 3 over F 2

n r f ( n-r ) () x g () r () x (1.1) = Σ g() x = Σ n f < -n+ r> g () r -n + r dx r dx n + ( -n,m) dx -n n+1 1 -n -1 + ( -n,n+1)

MATH 38061/MATH48061/MATH68061: MULTIVARIATE STATISTICS Solutions to Problems on Matrix Algebra

1. Matrix Algebra and Linear Economic Models

Endogoneity and All That

M in ing the Com pa tib ility Law of M ultid im en siona l M ed ic ines Ba sed on D ependence M ode Sets

A study on generalized absolute summability factors for a triangular matrix

Exercises 10. Find a fundamental matrix of the given system of equations. Also find the fundamental matrix Φ(t) satisfying Φ(0) = I. 1.

HOSVD. Higher Order Data Classification Method with Autocorrelation Matrix Correcting on HOSVD. Junichi MORIGAKI and Kaoru KATAYAMA

Journal of Central South University (Science and Technology) Jun i p i q

Diane Hu LDA for Audio Music April 12, 2010

Resurvey of Possible Seismic Fissures in the Old-Edo River in Tokyo

FREE VIBRATION OF A SINGLE-DEGREE-OF-FREEDOM SYSTEM Revision B

CHAPTER 103 EVEN AND ODD FUNCTIONS AND HALF-RANGE FOURIER SERIES

Ψηφιακό Μουσείο Ελληνικής Προφορικής Ιστορίας: πώς ένας βιωματικός θησαυρός γίνεται ερευνητικό και εκπαιδευτικό εργαλείο στα χέρια μαθητών

T he Op tim al L PM Po rtfo lio M odel of H arlow s and Its So lving M ethod

J. of Math. (PRC) Banach, , X = N(T ) R(T + ), Y = R(T ) N(T + ). Vol. 37 ( 2017 ) No. 5

Solutions: Homework 3

«ΧΩΡΙΚΗ ΜΟΝΤΕΛΟΠΟΙΗΣΗ ΤΗΣ ΚΑΤΑΝΟΜΗΣ ΤΟΥ ΠΛΗΘΥΣΜΟΥ ΤΗΣ ΠΕΡΔΙΚΑΣ (ALECTORIS GRAECA) ΣΤΗ ΣΤΕΡΕΑ ΕΛΛΑΔΑ»

Introduction of Numerical Analysis #03 TAGAMI, Daisuke (IMI, Kyushu University)

ΙΠΛΩΜΑΤΙΚΗ ΕΡΓΑΣΙΑ. ΘΕΜΑ: «ιερεύνηση της σχέσης µεταξύ φωνηµικής επίγνωσης και ορθογραφικής δεξιότητας σε παιδιά προσχολικής ηλικίας»

HIV HIV HIV HIV AIDS 3 :.1 /-,**1 +332

Bundle Adjustment for 3-D Reconstruction: Implementation and Evaluation

ΕΥΡΕΣΗ ΤΟΥ ΔΙΑΝΥΣΜΑΤΟΣ ΘΕΣΗΣ ΚΙΝΟΥΜΕΝΟΥ ΡΟΜΠΟΤ ΜΕ ΜΟΝΟΦΘΑΛΜΟ ΣΥΣΤΗΜΑ ΟΡΑΣΗΣ

ΤΕΧΝΟΛΟΓΙΚΟ ΕΚΠΑΙΔΕΥΤΙΚΟ ΙΔΡΥΜΑ ΗΡΑΚΛΕΙΟ ΚΡΗΤΗΣ ΣΧΟΛΗ ΔΙΟΙΚΗΣΗΣ ΚΑΙ ΟΙΚΟΝΟΜΙΑΣ ΤΜΗΜΑ ΛΟΓΙΣΤΙΚΗΣ

EE 570: Location and Navigation

Presentation of complex number in Cartesian and polar coordinate system

«ΙΕΡΕΥΝΗΣΗ ΤΩΝ ΠΑΡΑΓΟΝΤΩΝ ΠΟΥ ΕΠΙ ΡΟΥΝ ΣΤΗΝ ΑΦΟΣΙΩΣΗ ΤΟΥ ΠΕΛΑΤΗ ΣΕ ΕΠΩΝΥΜΑ ΠΡΟΪΟΝΤΑ ΤΡΟΦΙΜΩΝ. Η ΠΕΡΙΠΤΩΣΗ ΤΩΝ ΕΠΩΝΥΜΩΝ ΓΑΛΑΚΤΟΚΟΜΙΚΩΝ ΠΡΟΪΟΝΤΩΝ»

Inverse trigonometric functions & General Solution of Trigonometric Equations

COMMON RANDOM FIXED POINT THEOREMS IN SYMMETRIC SPACES

ΒΙΟΜΗΧΑΝΙΚΑ ΗΛΕΚΤΡΟΝΙΚΑ ΤΕΧΝΗΤΗ ΟΡΑΣΗ

On Certain Subclass of λ-bazilevič Functions of Type α + iµ

ΑΝΤΙΚΕΙΜΕΝΟ ΘΕΜΑ 1 Ο

Πανεπιστήμιο Κρήτης, Τμήμα Επιστήμης Υπολογιστών Άνοιξη HΥ463 - Συστήματα Ανάκτησης Πληροφοριών Information Retrieval (IR) Systems

w o = R 1 p. (1) R = p =. = 1

PACS: Mj, Dp, Ta

SIMULTANEOUS HYPOTHESIS TESTING OF SPLINE TRUNCATED MODEL IN NONLINEAR STRUCTURAL EQUATION MODELING (SEM)

ΣΤΗ ΣΧΕ ΙΑΣΗ ΕΚΠΑΙ ΕΥΤΙΚΟΥ ΛΟΓΙΣΜΙΚΟΥ 1

The Equivalence Theorem in Optimal Design

Jordan Form of a Square Matrix

Bayesian statistics. DS GA 1002 Probability and Statistics for Data Science.

Fourier Series. Fourier Series

Exercises to Statistics of Material Fatigue No. 5

:,,,,,, Engle 1982 ARCH,,,, (Realized Volatility),, Andersen Bollerslev Diebold Ebens(1998,2001), ( ) :

Πτυχιακή Εργασι α «Εκτι μήσή τής ποιο τήτας εικο νων με τήν χρή σή τεχνήτων νευρωνικων δικτυ ων»

Ανάκτηση Εικόνας βάσει Υφής με χρήση Eye Tracker

ΠΕΡΙΛΗΨΗ. Λέξεις κλειδιά: Υγεία και συμπεριφορές υγείας, χρήση, ψυχότροπες ουσίες, κοινωνικό κεφάλαιο.

ΠΤΥΧΙΑΚΗ ΕΡΓΑΣΙΑ "ΠΟΛΥΚΡΙΤΗΡΙΑ ΣΥΣΤΗΜΑΤΑ ΛΗΨΗΣ ΑΠΟΦΑΣΕΩΝ. Η ΠΕΡΙΠΤΩΣΗ ΤΗΣ ΕΠΙΛΟΓΗΣ ΑΣΦΑΛΙΣΤΗΡΙΟΥ ΣΥΜΒΟΛΑΙΟΥ ΥΓΕΙΑΣ "

«ΑΝΑΠΣΤΞΖ ΓΠ ΚΑΗ ΥΩΡΗΚΖ ΑΝΑΛΤΖ ΜΔΣΔΩΡΟΛΟΓΗΚΩΝ ΓΔΓΟΜΔΝΩΝ ΣΟΝ ΔΛΛΑΓΗΚΟ ΥΩΡΟ»

IPSJ SIG Technical Report Vol.2014-CE-127 No /12/6 CS Activity 1,a) CS Computer Science Activity Activity Actvity Activity Dining Eight-He

«ΑΓΡΟΤΟΥΡΙΣΜΟΣ ΚΑΙ ΤΟΠΙΚΗ ΑΝΑΠΤΥΞΗ: Ο ΡΟΛΟΣ ΤΩΝ ΝΕΩΝ ΤΕΧΝΟΛΟΓΙΩΝ ΣΤΗΝ ΠΡΟΩΘΗΣΗ ΤΩΝ ΓΥΝΑΙΚΕΙΩΝ ΣΥΝΕΤΑΙΡΙΣΜΩΝ»

IIT JEE (2013) (Trigonomtery 1) Solutions

A Decomposition Algorithm for the Solution of Fractional Quadratic Riccati Differential Equations with Caputo Derivatives

ST5224: Advanced Statistical Theory II

Automatic extraction of bibliography with machine learning

Gauss Radau formulae for Jacobi and Laguerre weight functions

0 irotmttm * eka.ia.gtxi Me ΤΗΝ ΠΡΟΝΟΙλ ΜΗΤΡΟΠΟΛΙΤΟΥ Μ V ΤI \ Η Ν Η C

Areas and Lengths in Polar Coordinates

The measurement of similarity in stock data documents collections

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

Tired Waiting in Queues? Then get in line now to learn more about Queuing!

, -.

Αντιλήψεις εκπαιδευτικών για τα περιβαλλοντικά ζητήµατα και σχολικά βιβλία: Η περίπτωση του φαινοµένου του θερµοκηπίου και του στρώµατος του όζοντος

Other Test Constructions: Likelihood Ratio & Bayes Tests

ΔΘΝΙΚΗ ΥΟΛΗ ΓΗΜΟΙΑ ΓΙΟΙΚΗΗ ΙΗ ΔΚΠΑΙΓΔΤΣΙΚΗ ΔΙΡΑ

ANOTHER EXTENSION OF VAN DER CORPUT S INEQUALITY. Gabriel STAN 1

DiracDelta. Notations. Primary definition. Specific values. General characteristics. Traditional name. Traditional notation

Retrieval of Seismic Data Recorded on Open-reel-type Magnetic Tapes (MT) by Using Existing Devices

Last Lecture. Biostatistics Statistical Inference Lecture 19 Likelihood Ratio Test. Example of Hypothesis Testing.

Applying Markov Decision Processes to Role-playing Game

Buried Markov Model Pairwise

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Wiki. Wiki. Analysis of user activity of closed Wiki used by small groups

42. διαβάζει την εφηµερίδα (α) ή να διαβάζει την εφηµερίδα (β) ii) Ορίζουµε το ενδεχόµενο

Current Status and Future Prospects of Camera-Based Character Recognition and Document Image Analysis

( ) 2 and compare to M.

The Spiral of Theodorus, Numerical Analysis, and Special Functions

Transcript:

(MIRU9) 9 7 R R R 1 R, R,..., R R Bisser Raytchev 739-857 1-4-1 63-19 8916 5 E-mail: {tamaki,bisser,ki}@hiroshima-u.ac.jp, amao@is.aist.jp 3 3 R R, R 3,..., R R R, R 3,... 3 3 R, R,..., R R R R,,, Abstract Ca R estimate a rotatio matrix R more accurately tha R? A method for estimatig a rotatio matrix R by usig R, R, R 3,... obtaied by a oe-shot measuremet Toru TAMAKI, Bisser RAYTCHEV, Toshiyuki AMANO, ad Kazufumi KANEDA Departmet of Iformatio Egieerig, Graduate School of Egieerig, Hiroshima Uiversity 1-4-1 Kagamiyama, Higashi-hiroshima, Hiroshima, 739-857 Japa Graduate School of Iformatio Sciece, Nara Istitute of Sciece ad Techology 8916-5 Takayama, Ikoma, Nara, 63-19 Japa E-mail: {tamaki,bisser,ki}@hiroshima-u.ac.jp, amao@is.aist.jp I this paper, we show that a more accurate estimatio of a 3 3 rotatio matrix R ca be achieved by appropriately decomposig higher-order rotatio matrices: R, R 3, ad so o. First we discuss a agle estimatio of a rotatio matrix ispired by the Electroic Distace Measuremet. The we reformulate the problem for a 3 3 rotatio matrix: if oise-cotamiated measuremet matrices R, R,..., R are give, fid a appropriate rotatio matrix R. I the proposed method, the give measuremet matrices are first trasformed to rotatio matrices by usig the polar decompositio. The the rotatio agles are obtaied by usig a eige decompositio of the rotatio matrices. Fially, the ambiguity of the obtaied rotatio agle is removed. Numerical simulatios ad pose estimatio experimets show that the use of R results i more accurate estimates tha whe R itself is used. Key words pose estimatio Rotatio matrix, Measuremet with higher frequecy, Electroic distace measuremet, View-based 1. R t 3

[1], [] F E R t [3] [1], [] ICP [4] [5] [6] [8] [9], [1] R R SO(3) [11] [14] R R R, R 3,... (Electroic Distace Measuremet, EDM) [15], [16] 3 3 R, R 3,... R R, R,..., R R 8 [] 8 1 R, R,..., R [17] 1 R R [6] [8] R R, R 3,... R, R, R 3,... 3 3 3 4 3 R R. 1 R R R 1 R θ R R θ R R R. 1 [16], [18] 1 Dist λ 1 1 θ 1 λ 1 > Dist Dist = θ 1 π λ 1, < = θ 1 < π (1)

1 Dist λ 1 λ k 1 k θ θ π θ = θ + π θ θ + π θ θ + πk θ 1 ( ) ( ) cos θ 1 cos( θ mi + πk ) k =,1 si θ 1 si( θ. (5) + πk ) θ 1 π/1 Dist λ 1 /1 λ (< λ 1 ) θ ( ) θ Dist = π + k λ, < = θ < π () k λ λ λ 1 1 1 1 1 λ /1 k Total Statio Geodimeter Tellurometer [19] (beatig). R R R R θ R 1 cos θ 1, si θ 1 R cos θ, si θ 1 θ θ 1 θ θ θ R R θ 1 θ θ θ 1 1 < = θ < π λ 1 = π θ θ θ 1 λ = π θ = θ 1 π λ 1 = θ 1, (3) ( ) θ θ = π + k λ = θ + πk, (4) θ 1 θ + πk R R θ R θ θ θ θ λ = π/ ( ) θ θ = π + k λ = θ + π k, (6) k =, 1,..., 1 k ( ) ( ) cos θ 1 cos( ˆk θ = argmi + π k ) k si θ 1 si( θ + π k ) θ + π ˆk 3. 3 3 (7) R 3 3 R 3 1 3 R ( = 1,,...) 3 3 3. 1 R 3 3 R R R R R R [], [1] mi R R F + µ R T R I F (8) µ

polar decompositio [] R R = U Σ V T R polar decompositio R = (U V T )(V Σ V T ). (9) polar part U V T (8) U V T 1 U V T R { U V R T if U V = +1, = (1) U (HV ) T if U V = 1, H = diag(1, 1, 1) polar decompositio R U V T V T V I G V T G 1 GV polar part 1 H V T H 1 HV H T H = I H = 1 U V = 1 polar decompositio R = (U (HV ) T )(HV Σ V T ) (11) 3. R R expoetial map [] cos θ = trr 1 expoetial map π θ π cos si ω = (b, c, d) T 3 expoetial map [] ω = (r 3 r 3, r 13 r 31, r 1 r 1 ) T, (1) r ij R ij R R 1 [1], [] R P D [1], [], [3] R = P D P H, (13) i = 1 b P = 1 c ω d c +d bc+id ω (c +d ) bd ic ω (c +d ) c +d bc id ω (c +d ) bd+ic ω (c +d ) (14) D = diag(1, e iθ, e iθ ) (15) P 1 D e ±iθ R P = D θ P H P R R = P DP H R P H P = I R = P DP H P DP H = P D P H P R P ω 1,..., ω media 1 1 ω med P med ω med = med ω i (16) i=1,..., media media 3. 3 [4], [5] Matlab S S = 1 1 i 1 i, S H S = SS H = I. (17) R = P D P H = P SS H D SS H P H, (18) = P D P T, (19) P = P S D = S H D S P 1 media

P P T = P T P = I D D = 1 cos θ si θ si θ cos θ. () θ D = P T R P 3. 4 R 1 R θ 1 θ θ ( ) ( ) cos θ 1 cos( ˆk θ = argmi + π k ) k=,1, si θ 1 si( θ..., 1 + π k ) ˆθ = θ + π ˆk (1) R 1 R θ 1 θ 1 θ 1 π k θ 1 θ 1 π 1 k 1 R = P med cos ˆθ si ˆθ P T si ˆθ cos ˆθ med () 4. R ( = 1,,...) R 1 (R 1 ) R 4. 1 1 R R R R R R [6] 1 4 agle error [deg] F orm 3 F orm 4 15 1 5.1..3.4.5 1 3 4 5 6 7 8 1 ±.1 ±.5 1.8.6.4..1..3.4.5 1 3 4 5 6 7 8 1 ±.1 ±.5 1.8.6.4..1..3.4.5 1 3 4 5 6 7 8 1 ±.1 ±.5 P med P 1 1 R 1 R ±.1 ±.5

θ, ω ˆθ, ˆω 3 θ, ω (π θ), ω ω ˆω 9 π θ ˆθ, ω T ˆω < π agle error = θ (π ˆθ), (3) otherwise agle error [deg] 15 1 5.1..3.4.5 [deg] R = 1 R 1 > 1 R 1 R,..., R 1 (±.1,.,.3,.4,.5) R, R 3,... R 1 R R 1 7 3 R R 1 R 3 R media P med media R P med P 1 4 R R 1 R 3 P P 4. R R R 1 R 1 (R 1 ) R 1 3 4 5 6 7 8 5 ±.1 ±.5 F orm 1.1..3.4.5.8.6.4. 1 3 4 5 6 7 8 6 ±.1 ±.5 R 1 5 6 R R 1 R R 1 R R 1, R,... R 1 4. 3 3 [6] [8] [5] [9], [1] 3 R R, R,..., R 8

agle error [deg] 1 1 8 6 4 1 3 4 5 6 7 8 7 F orm.4.3..1 1 3 4 5 6 7 8 8 R 1,..., R 8 [8] [7] [8] R x R R = F x R, = 1,..., 8 (4) F i R R 9 x R = F x, = 1,..., 8 (5) 9 R 3 3 R 3 1 1 CG 18 18 3 1 3 3[deg] 7 8 R 1 5.31 ± 5.84[deg] R 8.87 ± 4.55[deg] R 1.183 ±.16 R 8.14 ±.146 5. R R ( > 1) R [1] Richard Hartley ad Adrew Zisserma. Multiple View Geometry i Computer Visio. Cambridge Uiversity Press, d editio, 4. http://www.cambridge.org/aus/catalogue/ catalogue.asp?isb=5154518. [] Yi Ma, Stefao Soatto, Jaa Košecká, ad S. Shakar Sastry. A Ivitatio To 3-D Visio. Spriger, 4. http://visio.ucla.edu/masks/. [3] Carlo Tomasi ad Takeo Kaade. Shape ad motio from image streams uder orthography: a factorizatio method. Itl. J. of Computer Visio, Vol. 9, No., pp. 137 154, 199. http://www. sprigerlik.com/cotet/q3546r136334l8r. [4] P. J. Besl ad N. D. McKay. A method for registratio of 3-D shapes. IEEE Tras. PAMI, Vol. 14, No., pp. 39 56, 199. http://doi. ieeecomputersociety.org/1.119/34.11791. [5] D. G. Lowe. Fittig parameterized three-dimesioal models to images. IEEE Tras. PAMI, Vol. 13, No. 5, pp. 441 45, 1991. http://doi. ieeecomputersociety.org/1.119/34.13443. [6] Hiroshi Murase ad Shree K. Nayar. Visual learig ad recogitio of 3-D objects from appearace. Itl. J. of Computer Visio, Vol. 14, No. 1, pp. 5 4, 1995. http://dx.doi.org/1.17/bf141486. [7] Gabriele Peters, Barbara Zitova, ad Christoph

vo der Malsburg. How to measure the pose robustess of object views. Image ad Visio Computig, Vol., No. 4, pp. 49 56,. http://ls7-www.cs.ui-dortmud.de/ peters/ pages/research/modeladaptsys/ modeladaptsys vba rov.html. [8] Thomas Melzer, Michael Reiter, ad Horst Bischof. Appearace models based o kerel caoical correlatio aalysis. Patter Recogitio, Vol. 36, pp. 1961 1971, 3. http: //dx.doi.org/1.116/s31-33(3)58-x. [9] David G. Lowe. Distictive image features from scale-ivariat keypoits. Itl. J. of Computer Visio, Vol. 6, No., pp. 91 11, 4. http://www.sprigerlik.com/cotet/ h4l69137px768. [1] Fred Rothgager, Svetlaa Lazebik, Cordelia Schmid, ad Jea Poce. 3D object modelig ad recogitio usig affie-ivariat patches ad multi-view spatial costraits. Proc. of CVPR3, Vol., pp. 7 77, 3. http://www-cvr.ai.uiuc.edu/poce grp/ publicatio/paper/cvpr3a.ps.gz. [11] Ameesh Makadia ad Kostas Daiilidis. Rotatio recovery from spherical images without correspodeces. IEEE Tras. PAMI, Vol. 8, No. 7, pp. 117 1175, 6. [1] Ameesh Makadia ad Kostas Daiilidis. Direct 3d-rotatio estimatio from spherical images via a geeralized shift theorem. Proc. of CVPR3, Vol., p. 17, 3. http://doi.ieeecomputersociety.org/1.119/ CVPR.3.111473. [13] Peter J. Kostelec ad Daiel N. Rockmore. Ffts o the rotatio group. Joural of Fourier Aalysis ad Applicatios, Vol. 14, No., 8. http://www. sprigerlik.com/cotet/r631816xr89vr5. [14] David K. Masle ad Daiel M. Rockmore. Geeralized FFTs a survey of some recet results. Groups ad Computatio II, DIMACS Series i Discrete Mathematics ad Theoretical Computer Sciece, pp. 183 38, 1997. http://www.cs. dartmouth.edu/ rockmore/dimacs-.pdf. [15].., 1976. [16] Russell Charles Briker ad Roy Miick, editors. The Surveyig Hadbook. Chapma & Hall, d editio, 1995. http: //books.google.com/books?id=gb7w9xlnjac. [17] Zhegyou Zhag. Parameter estimatio techiques: a tutorial with applicatio to coic fittig. Image ad Visio Computig, Vol. 15, No. 1, pp. 59 76, 1997. [18],,,,.., 8. http: //books.google.co.jp/books?id=jk7ivyjsloc. [19] The Alberta Lad Surveyors Associatio. Equipmet. olie, accessed 9//6. http:// www.ladsurveyighistory.ab.ca/equipmet.htm. [] Gee Howard Golub ad Charles F. Va Loa. Matrix Computatios. The Joh Hopkis Uiversity Press, 3rd, 1996. http: //books.google.com/books?id=mloa7wpx6oyc. [1] Zhegyou Zhag. A flexible ew techique for camera calibratio. Techical Report MSR-TR-98-71, Microsoft Research, 1998. http://research.microsoft.com/research/pubs/ view.aspx?tr id=1. [] MathPages. Rotatio matrices. olie, accessed 9//6. http://www.mathpages.com/home/ kmath593/kmath593.htm. [3] Eric W. Weisstei. Rotatio matrix. From MathWorld A Wolfram Web Resource, accessed 9//6. http: //mathworld.wolfram.com/rotatiomatrix.html. [4] Ichiro Satake. Liear Algebra. Marcel Dekker Ic., 1975. [5],,.. MIRU7, pp. 958 963, 7. http://ir.lib.hiroshima-u.ac.jp/1989. [6] GSL GNU Scietific Library, 8. http://www.gu.org/software/gsl/. [7] Shigo Ado, Yoshiori Kusachi, Akira Suzuki, ad Keichi Arakawa. Appearace based pose estimatio of 3D object usig support vector regressio. ICIP5, Vol. 1, pp. I 341 344, 5. http://ieeexplore.ieee.org/xpls/abs all.jsp? arumber=159757&isumber=366. [8] Takayuki Okatai ad Koichiro Deguchi. Yet aother appearace-based method for pose estimatio based o a liear model. IAPR Workshop o Machie Visio Applicatios, pp. 58 61,. http: //b.cvl.iis.u-tokyo.ac.jp/mva/proceedigs/ CommemorativeDVD//papers/58.pdf.