Index. 3D reconstruction, point algorithm, point algorithm, point algorithm, point algorithm, 263

Similar documents
Index. 3D reconstruction, point algorithm, point algorithm, point algorithm, point algorithm, 253

calibrated coordinates Linear transformation pixel coordinates

Unit 3 Multiple View Geometry

55:148 Digital Image Processing Chapter 11 3D Vision, Geometry

Stereo and Epipolar geometry

Multiple View Geometry in Computer Vision Second Edition

Part I: Single and Two View Geometry Internal camera parameters

Structure from Motion. Introduction to Computer Vision CSE 152 Lecture 10

Outline. ETN-FPI Training School on Plenoptic Sensing

Multiple View Geometry in Computer Vision

Week 2: Two-View Geometry. Padua Summer 08 Frank Dellaert

Parameter estimation. Christiano Gava Gabriele Bleser

Computer Vision Projective Geometry and Calibration. Pinhole cameras

Agenda. Rotations. Camera models. Camera calibration. Homographies

Camera calibration. Robotic vision. Ville Kyrki

Vision Review: Image Formation. Course web page:

Structure from motion

Rigid Body Motion and Image Formation. Jana Kosecka, CS 482

Contents. 1 Introduction Background Organization Features... 7

Rectification and Distortion Correction

Two-view geometry Computer Vision Spring 2018, Lecture 10

CS201 Computer Vision Camera Geometry

Structure from motion

Multiple Views Geometry

Computational Optical Imaging - Optique Numerique. -- Single and Multiple View Geometry, Stereo matching --

Camera Calibration. Schedule. Jesus J Caban. Note: You have until next Monday to let me know. ! Today:! Camera calibration

Lecture 9: Epipolar Geometry

Epipolar Geometry Prof. D. Stricker. With slides from A. Zisserman, S. Lazebnik, Seitz

Vision par ordinateur

Chapter 7: Computation of the Camera Matrix P

CS 231A: Computer Vision (Winter 2018) Problem Set 2

Hartley - Zisserman reading club. Part I: Hartley and Zisserman Appendix 6: Part II: Zhengyou Zhang: Presented by Daniel Fontijne

Multiview Stereo COSC450. Lecture 8

Agenda. Rotations. Camera calibration. Homography. Ransac

N-Views (1) Homographies and Projection

Reminder: Lecture 20: The Eight-Point Algorithm. Essential/Fundamental Matrix. E/F Matrix Summary. Computing F. Computing F from Point Matches

COMPARATIVE STUDY OF DIFFERENT APPROACHES FOR EFFICIENT RECTIFICATION UNDER GENERAL MOTION

CS231A Course Notes 4: Stereo Systems and Structure from Motion

Computational Optical Imaging - Optique Numerique. -- Multiple View Geometry and Stereo --

Stereo Vision. MAN-522 Computer Vision

Rectification and Disparity

55:148 Digital Image Processing Chapter 11 3D Vision, Geometry

arxiv: v1 [cs.cv] 28 Sep 2018

1 Projective Geometry

Stereo II CSE 576. Ali Farhadi. Several slides from Larry Zitnick and Steve Seitz

Step-by-Step Model Buidling

Computer Vision Projective Geometry and Calibration. Pinhole cameras

Epipolar Geometry and Stereo Vision

Dense 3D Reconstruction. Christiano Gava

CSE 252B: Computer Vision II

Homogeneous Coordinates. Lecture18: Camera Models. Representation of Line and Point in 2D. Cross Product. Overall scaling is NOT important.

Camera model and multiple view geometry

BIL Computer Vision Apr 16, 2014

Homographies and RANSAC

Geometric camera models and calibration

Today. Stereo (two view) reconstruction. Multiview geometry. Today. Multiview geometry. Computational Photography

Geometry for Computer Vision

Lecture 14: Basic Multi-View Geometry

Cameras and Stereo CSE 455. Linda Shapiro

CS231A Midterm Review. Friday 5/6/2016

Camera models and calibration

CS 664 Structure and Motion. Daniel Huttenlocher

Today s lecture. Image Alignment and Stitching. Readings. Motion models

Two-View Geometry (Course 23, Lecture D)

Machine vision. Summary # 11: Stereo vision and epipolar geometry. u l = λx. v l = λy

Epipolar Geometry and the Essential Matrix

Camera Geometry II. COS 429 Princeton University

CS231M Mobile Computer Vision Structure from motion

CS 532: 3D Computer Vision 7 th Set of Notes

Application questions. Theoretical questions

CSE 252B: Computer Vision II

Epipolar Geometry and Stereo Vision

Advanced Computer Graphics Transformations. Matthias Teschner

3D Geometry and Camera Calibration

Dense 3D Reconstruction. Christiano Gava

3D Modeling using multiple images Exam January 2008

Image Rectification (Stereo) (New book: 7.2.1, old book: 11.1)

Lecture 5 Epipolar Geometry

Final Exam Study Guide CSE/EE 486 Fall 2007

CSCI 5980/8980: Assignment #4. Fundamental Matrix

Multiple View Geometry in Computer Vision

CSE 252B: Computer Vision II

Last lecture. Passive Stereo Spacetime Stereo

But First: Multi-View Projective Geometry

Compositing a bird's eye view mosaic

Visual Recognition: Image Formation

Epipolar Geometry in Stereo, Motion and Object Recognition

Computer Vision Lecture 17

Computer Vision Lecture 17

Multiple View Geometry. Frank Dellaert

Recovering structure from a single view Pinhole perspective projection

Structure from Motion

Epipolar geometry. x x

CS6670: Computer Vision

3D Computer Vision. Structure from Motion. Prof. Didier Stricker

CS 231A Computer Vision (Winter 2015) Problem Set 2

C280, Computer Vision

3-D D Euclidean Space - Vectors

Structure from Motion and Multi- view Geometry. Last lecture

Image Stitching. Slides from Rick Szeliski, Steve Seitz, Derek Hoiem, Ira Kemelmacher, Ali Farhadi

Transcription:

Index 3D reconstruction, 125 5+1-point algorithm, 284 5-point algorithm, 270 7-point algorithm, 265 8-point algorithm, 263 affine point, 45 affine transformation, 57 affine transformation group, 57 affine vector, 45 algebraic distance, 39 algebraic error, 39, 188 algebraic errors, 188 algebraic representation, 147 algebraic representation of SO(3), 149 algebraically independent, 120 analytic geometry, 16 aperture, 95, 292 associative operation, 348 automatic camera calibration, 290 auxiliary variables, 205, 228, 268, 303 axis-angle representation, 150 backward interpolation, 311 barrel distortion, 295 barycentric coordinates, 246 baseline, 125 blending, 313 block diagonal matrix, 356 block matrix, 355 bookkeeping tables, 330 bounding box, 110, 306 Brown-Conrady distortion, 297 bundle adjustment, 17, 328 incremental, 328 C-normalized image coordinates, 100 calibrated epipolar geometry, 123, 262 calibration device, 300 camera, 95 camera calibration, 110, 224, 289 camera center, 94 camera centered coordinate systems, 96 camera matrices from F, 132 camera matrix, 94 camera normalized image coordinates, 100 camera obscura, 95 camera projection matrix, 94 camera resectioning, 108 canonical form, 30, 32, 65, 66, 70, 73 canonical homogeneous coordinates, 90 Cardano angles, 162 Cartesian coordinate system, 21 Cartesian representation, 22 Cayley transform, 156 central perspective, 93 central projection, 93 centroid, 175 Closed form concatenation of rotations, 150 Closed form inverse of a rotation, 150 co-linear planes, 26 co-linear points, 22, 26 co-planar lines, 26 co-planar points, 25 collineation, 85 column compression, 360 commutative operation, 348 complete representation, 149 consensus set, 275 consistency, 190 consistent parameterization, 228 consistent Plücker coordinates of 3D line, 72 constraint enforcement, 229 fundamental matrix, 263 normalized camera matrix, 252 rotation matrix, 243 corner points, 179 correspondence, 123 correspondence problem, 206, 207 corresponding points, 123, 203 cost function, 174 cross product, 351 cross product matrix, 351 cross product operator, 351 cross-correlation, 175 D-normalization, 33 data constraint, 190 data degeneracy, 191, 209 data error, 180 data matrix, 36, 189 decentering distortion, 297 390

degeneracy epipolar geometry estimation, 266 homography estimation, 209 method, 194 of data, 191 degenerate configuration, 52 degenerate data, 35 degrees of freedom, 51, 190, 227 depth, 99 depth-of-field, 292 descent direction, 376 direct linear transformation, 119, 205 directed line, 41 direction, 45 distibutive operation, 348 distortion center, 298 DL-normalization, 73 DLT, 205 double embedding of SO(3), 159 dual homogeneous coordinates, 32, 67 dual line normalization, 33, 73 dual Plücker coordinates, 49, 71 dual transformation, 62 duality, 50 duality mapping, 72 Eckart-Young (-Mirsky) teorem, 364 ego-motion estimation, 328 energy function, 174 epipolar constraint, 120, 123, 128 epipolar geometry, 123 degenerate, 266 epipolar line, 126, 129 epipolar plane, 126 epipolar points, 125 epipoles, 125 equation of the line, 22 equation of the plane, 25 equilibrating transformation, 219 equivalent cameras, 115 error function, 174 essential matrix, 123, 141 estimation robust, 284 internal constraint, 142 estimation camera matrix, 223 epipolar geometry, 255 fundamental matrix, 262 homography, 203 introduction, 173 MAP, 199 maximum a posteriori probability, 199 rigid transformation, 245 rotations, 241 transforamtions, 203 Euclidean geometry, 15 Euclidean reconstruction, 136 Euler angles, 162 exterior orientation, 107, 245 external camera parameters, 107 feature point, 179 FET, 296 field of view, 121 field of view model, 296 fish-eye lens, 292 fish-eye transform, 296 focal length, 96 focal point, 94 forward interpolation, 311 FOV, 296 full rank pseudo-inverse, 353 fundamental matrix, 123, 128 camera matrices from, 132 estimation, 262 robust, 283 internal constraint, 128 Gauss-Newton method, 380 general configuration, 52 geometric distortion, 292 geometric objects, 29 gimbal lock, 169 gold standard method, 268 H-normalization, 219 hand-eye calibration, 245 Hartley normalization, 219 homogeneous coordinates, 17, 28, 30, 65 homogeneous method, 196 homogeneous representation, 22 homography, 85 homography estimation degeneracy, 209 robust, 282 homography group, 87 homography transformation, 85 hybrid method, 380 ICP, 247 ideal data, 180 ideal point, 16, 45 identity element, 348 identity rotation, 150 IID, 181 image coordinate system, 103 image coordinates, 94 image mosaic, 306 391

image mosaicking, 306 image plane, 94 image point, 94 incidence relations, 34 incremental bundle adjustment, 328 independent and identical distributions, 181 indeterministic estimation method, 275 inhomogeneous method, 193 inliers, 183, 272 inner derivatives, 230 intercept, 24 interest point, 179 interior orientation, 107 internal camera parameters, 107 internal constraint, 190 essential matrix, 142 fundamental matrix, 128 normalized camera matrix, 101 of a rotation representation, 150 Plücker coordinates, 72 rotation matrix, 148 invariances, 51 inverse Cayley transform, 157 inverse of a quaternion, 350 involution, 59 iterative closest point, 247 Jacobian, 379 Jacobian mask, 340, 384 keystone effect, 88 L-normalization, 70 L 1 error, 185 L 2 error, 185 left inverse, 354 left singular vectors, 357 lens distortion, 292 lens distortion function, 293 lens effect, 292 Levenberg-Marquardt method, 380 likelihood, 200 line at infinity, 46, 69 line normalization, 70 linear transformation, 347 MAP, 199 matrix exponential function, 366 matrix logarithm, 367 max-error, 186 maximum a posteriori probability, 199 Maximum Likelihood estimation, 200 maximum likelihood estimation, 200 mean, 175 measurement error, 180 measurement noise, 113, 180 method degeneracies, 261 method degeneracy, 194, 209 mid-point method, 256 minimal case estimation, 208 minimal case of model estimation, 191 minimal parameterization, 106, 132 ML-estimation, 200 model error, 181 model estimation minimal case, 191 over-determined case, 192 under-determined case, 191 model parameters, 175, 189 Moore-Penrose inverses, 354 motion stereo, 125 nearest neighbor interpolation, 113, 312 Newton s method, 378 non-linear optimization, 184, 375 normalization constraint, 190 normalized 8-point algorithm, 264 normalized camera, 99 normalized camera matrix, 100, 101 internal constraint, 101 normalizing transformation, 216, 219 objective function, 174 observation, 179, 180 omni-directional camera, 306 one-first enumeration, 112 OPP, 362 optical axis, 96 optical center, 94 optical ray, 114 optimal triangulation, 257 orientation of a line, 44 orthogonal Procrustes problem, 362 outlier, 187 outliers, 183, 272 over-determined model estimation, 192 over-fitting, 182 P-normalization, 31 P3P, 254 panorama image, 307 panorama stitching, 307 parameter space, 149 parameterization of a transformation, 52 parametric representation, 23 perspective n-point problem, 251 photogrammetry, 17 pin-cushion distortion, 295 pinhole camera, 93, 95 pinhole perspective, 93 392

pixel coordinates, 103 pixel density, 107 pixel grid, 111 pixel value, 111 pixels, 95 Plücker coordinates of 2D line, 48 Plücker coordinates of 3D line, 69 consistent, 72 internal constraint, 72 plumb bob distortion, 297 PnP, 251 point at infinity, 16, 44 point cloud, 336 point normalization, 31 points of interest, 179, 207 pose, 107 pre-conditioning, 219 principal axis, 96 principal line, 96 principal plane, 96 principal point, 96 probability density, 197 probability density function, 180 projection, 85 projection line, 85, 114 projection plane, 119 projection point, 85 projective elements, 346 projective geometry, 16, 46 projective group, 87 projective plane, 16, 46 projective reconstruction, 137 projective space, 346 projective spaces, 17 projectivity, 85 proper line, 69 proper lines, 42 proper points, 42 pseudo-inverse, 353 full rank, 353 pure quaternion, 350 putative correspondences, 324 quaternion, 348 inverse, 350 pure, 350 quaternion sandwich product, 158 quaternionic embedding of R 3, 158 quaternionic embedding of SO(3), 159 radial distortion function, 294 Random Sample Consensus, 276 RANSAC, 276 re-mapping of the residual vector, 231 re-normalize, 215 re-parameterization, 229 reconstruction, 135 rectified stereo rig, 125, 138 rectifying homographies, 139 reference image, 308 registration, 247 representative of a projective element, 345 reprojection error, 327 residual, 189, 379 residual error, 175 residual vector, 226 right inverse, 354 right singular vectors, 357 rigid transformation, 55 estimation, 245 rigid transformation group, 55 robust errors, 273 robust estimation, 271, 272 essential matrix, 284 fundamental matrix, 283 homography, 282 Rodrigues rotation formula, 151, 152 rotation, 148 rotation matrix internal constraint, 148 row compression, 360 SAD, 185 sandwich product, 84 Schur complement, 356 Schur complement trick, 341 SE(2), 55 SE(3), 81 SfM, 325 shearing, 60 signed distance, 39 similarity reconstruction, 137 similarity transformations, 56 singular value decomposition, 356 singular values, 357 singularity, 150 skewing, 60 slope, 24 solution space, 190, 209 SOPP, 243 special Euclidean transformation group in E 2, 55 in E 3, 81 special orthogonal Procrustes problem, 243 special QR-factorization, 372 special SVD, 363 SSD, 185 SSVD, 363 standard Plücker coordinates, 71 393

step length, 376 stereo camera, 123 stereo rectification, 125 stereo rig, 123, 125 structure and motion, 325 structure from motion, 325 sum of absolute differences, 185 sum of square differences, 185 SVD, 356 SVD profile, 211 Tait-Bryan angles, 162 thin-prism distortion, 297 three-angle representation, 162 total least squares, 177 transfer, 130 of epipolar lines, 130 of points, 130 translation group, 54 trial set, 275 triangulation, 17, 120, 125, 135, 255 twisted pair, 166 twisted rotations, 143, 166 unbiased estimate, 221 uncalibrated epipolar geometry, 123, 132 under-determined model estimation, 191 unique representation, 149 unit quaternion, 350 unperturbed data, 180 variances, 175 vector product, 351 vectorization, 190, 206, 224 virtual image plane, 99 visibility function, 327 visual appearance, 280 visual odometry, 328 Wahba s problem, 242 world coordinate system, 100 zero-first enumeration, 112 394