More on single-view geometry class 10 Multiple View Geometry Comp 290-089 Marc Pollefeys Multiple View Geometry course schedule (subject to change) Jan. 7, 9 Intro & motivation Projective 2D Geometry Jan. 14, 16 (no class) Projective 2D Geometry Jan. 21, 23 Projective 3D Geometry (no class) Jan. 28, 30 Parameter Estimation Parameter Estimation Feb. 4, 6 Algorithm Evaluation Camera Models Feb. 11, 13 Camera Calibration Single View Geometry Feb. 18, 20 Epipolar Geometry 3D reconstruction Feb. 25, 27 Fund. Matrix Comp. Structure Comp. Planes & Homographies Trifocal Tensor Mar. 18, 20 Three View Reconstruction Multiple View Geometry Mar. 25, 27 MultipleView Reconstruction Bundle adjustment Apr. 1, 3 Auto-Calibration Papers Apr. 8, 10 Dynamic SfM Papers Apr. 15, 17 Cheirality Papers Apr. 22, 24 Duality Project Demos Mar. 4, 6 Single view geometry Camera model Camera calibration Single view geom. Gold Standard algorithm Objective Given n≥6 2D to 2D point correspondences {Xi↔xi’}, determine the Maximum Likelyhood Estimation of P Algorithm (i) Linear solution: ~ (a) Normalization: Xi UX i ~x i Tx i (b) DLT (ii) Minimization of geometric error: using the linear estimate as a starting point minimize the geometric error: ~ ~~ (iii) Denormalization: P T-1~ PU More Single-View Geometry P T CP Q cone • Projective cameras and planes, lines, conics and quadrics. PQ *PT C* • Camera calibration and vanishing points, calibrating conic and the IAC Action of projective camera on planes X X Y x PX p1p 2 p 3p 4 p1p 2 p 4 Y 0 1 1 The most general transformation that can occur between a scene plane and an image plane under perspective imaging is a plane projective transformation (affine camera-affine transformation) Action of projective camera on lines forward projection Xμ P(A μB) PA μPB a μb back-projection PTl T X lT PX Action of projective camera on conics back-projection to cone Q co P CP T x Cx X P CPX 0 T T T example: T T T K K Qco C K | 0 CK 0 0 0 0 Images of smooth surfaces The contour generator G is the set of points X on S at which rays are tangent to the surface. The corresponding apparent contour g is the set of points x which are the image of X, i.e. g is the image of G The contour generator G depends only on position of projection center, g depends also on rest of P Action of projective camera on quadrics back-projection to cone C PQ P * * T T Q* lT PQ*P T l 0 The plane of G for a quadric Q is camera center C is given by =QC (follows from pole-polar relation) The cone with vertex V and tangent to the quadric Q is Q CO (V T QV)Q - (QV)(QV) T QCO V 0 The importance of the camera center ~ P KR[I | C], P' K' R' [I | C] P' K' R' KR P -1 x' P' X K' R' KR PX K' R' KR x -1 -1 x' Hx with H K' R' KR -1 Moving the image plane (zooming) x K[I | 0]X -1 x' K'[I | 0]X K' K x kI -1 H K' K T 0 (1 k)~ x0 1 ~ kI (1 k) x0 kI K' T K T 1 0 0 x0 kA ~ kI 0 T K T 0 1 1 0 k f / f' (1 k)~ x0 A ~ x0 1 0T 1 Camera rotation x K[I | 0]X x' K[R | 0]X KRK -1x H KRK -1 conjugate rotation μ, μe i , μe i Synthetic view (i) Compute the homography that warps some a rectangle to the correct aspect ratio (ii) warp the image Planar homography mosaicing close-up: interlacing can be important problem! Planar homography mosaicing more examples Projective (reduced) notation X1 (1,0,0,0) T , X 2 (0,1,0,0) T , X 3 (0,0,1,0) T , X 4 (0,0,0,1) T x1 (1,0,0) T , x 2 (0,1,0) T , x 3 (0,0,1) T , x 4 (1,1,1) T a 0 0 d P 0 b 0 d 0 0 c d C (a 1 , b 1 , c 1 , d 1 )T Moving the camera center motion parallax epipolar line What does calibration give? x K[I | 0]d 0 d K 1x cos T T d1 d 2 d d d T 1 1 T 2 d2 x x1 (K -T K -1 )x 2 T 1 T (K -T K -1 )x1 x 2 (K -T K -1 )x 2 An image l defines a plane through the camera center with normal n=KTl measured in the camera’s Euclidean frame The image of the absolute conic ~ d x PX KR[I | C] KRd 0 mapping between p∞ to an image is given by the planar homogaphy x=Hd, with H=KR image of the absolute conic (IAC) ω KK (i) (ii) (iii) (iv) (v) T 1 K -T K -1 C H IAC depends only on intrinsics angle between two rays cos DIAC=w*=KKT w K (cholesky factorisation) image of circular points T CH1 T x x1 ωx 2 T 1 T ωx1 x 2 ωx 2 A simple calibration device (i) compute H for each square (corners (0,0),(1,0),(0,1),(1,1)) (ii) compute the imaged circular points H(1,±i,0)T (iii) fit a conic to 6 circular points (iv) compute K from w through cholesky factorization (= Zhang’s calibration method) Orthogonality = pole-polar w.r.t. IAC The calibrating conic 1 K 1 C K T 1 1 Vanishing points xλ PXλ PA λPD a λKd v lim x λ lim a λKd Kd λ v PX Kd λ Vanishing lines Orthogonality relation cos v v1 ωv2 0 T l1 ω*l 2 0 T T 1 T 1 v ωv2 T ωv1 v 2 ωv2 Calibration from vanishing points and lines Calibration from vanishing points and lines Next class: Two-view geometry Epipolar geometry