Mysteries of Parameterizing Camera Motion - Part 1

Similar documents
Pin Hole Cameras & Warp Functions

Pin Hole Cameras & Warp Functions

Fundamental Matrix & Structure from Motion

Agenda. Rotations. Camera models. Camera calibration. Homographies

Introduction to Computer Vision

Structure from Motion

Structure from motion

COSC579: Scene Geometry. Jeremy Bolton, PhD Assistant Teaching Professor

Fundamental Matrix & Structure from Motion

Structure from motion

Camera Model and Calibration

Geometric camera models and calibration

Camera calibration. Robotic vision. Ville Kyrki

Agenda. Rotations. Camera calibration. Homography. Ransac

BIL Computer Vision Apr 16, 2014

Autonomous Navigation for Flying Robots

Humanoid Robotics. Projective Geometry, Homogeneous Coordinates. (brief introduction) Maren Bennewitz

Agenda. Perspective projection. Rotations. Camera models

Planar homographies. Can we reconstruct another view from one image? vgg/projects/singleview/

Parameter estimation. Christiano Gava Gabriele Bleser

Augmented Reality II - Camera Calibration - Gudrun Klinker May 11, 2004

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

CS231A Course Notes 4: Stereo Systems and Structure from Motion

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

Structure from Motion

Camera Model and Calibration. Lecture-12

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

Camera Parameters and Calibration

Visual Recognition: Image Formation

Structure from Motion CSC 767

ECE Digital Image Processing and Introduction to Computer Vision. Outline

COS429: COMPUTER VISON CAMERAS AND PROJECTIONS (2 lectures)

Unit 3 Multiple View Geometry

Structure from motion

Instance-level recognition I. - Camera geometry and image alignment

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

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

Perspective projection and Transformations

Computer Vision in a Non-Rigid World

Computer Vision Projective Geometry and Calibration. Pinhole cameras

Animation. Keyframe animation. CS4620/5620: Lecture 30. Rigid motion: the simplest deformation. Controlling shape for animation

Vision Review: Image Formation. Course web page:

Humanoid Robotics. Least Squares. Maren Bennewitz

CEE598 - Visual Sensing for Civil Infrastructure Eng. & Mgmt.

Camera Geometry II. COS 429 Princeton University

calibrated coordinates Linear transformation pixel coordinates

Epipolar geometry. x x

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

Two-view geometry Computer Vision Spring 2018, Lecture 10

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

Structure from Motion

The Lucas & Kanade Algorithm

CSE 252B: Computer Vision II

arxiv: v1 [cs.cv] 23 Apr 2017

The end of affine cameras

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

An Overview of Matchmoving using Structure from Motion Methods

CS231A. Review for Problem Set 1. Saumitro Dasgupta

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

Robust Geometry Estimation from two Images

Image warping , , Computational Photography Fall 2017, Lecture 10

arxiv: v1 [cs.cv] 28 Sep 2018

Flexible Calibration of a Portable Structured Light System through Surface Plane

N-View Methods. Diana Mateus, Nassir Navab. Computer Aided Medical Procedures Technische Universität München. 3D Computer Vision II

Camera Models and Image Formation. Srikumar Ramalingam School of Computing University of Utah

CS 664 Structure and Motion. Daniel Huttenlocher

Compositing a bird's eye view mosaic

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

Lecture 9: Epipolar Geometry

Camera Models and Image Formation. Srikumar Ramalingam School of Computing University of Utah

MERGING POINT CLOUDS FROM MULTIPLE KINECTS. Nishant Rai 13th July, 2016 CARIS Lab University of British Columbia

Computer Vision I - Appearance-based Matching and Projective Geometry

Last lecture. Passive Stereo Spacetime Stereo

3D Geometry and Camera Calibration

Monocular Visual Odometry

CS231M Mobile Computer Vision Structure from motion

Rectification and Disparity

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

Efficient Interest Point Detectors & Features

Multiple Motion Scene Reconstruction from Uncalibrated Views

Stereo. 11/02/2012 CS129, Brown James Hays. Slides by Kristen Grauman

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

Identifying Car Model from Photographs

An idea which can be used once is a trick. If it can be used more than once it becomes a method

Stereo Image Rectification for Simple Panoramic Image Generation

Structure from Motion and Multi- view Geometry. Last lecture

Computer Vision I - Algorithms and Applications: Multi-View 3D reconstruction

Recovering structure from a single view Pinhole perspective projection

Outline. ETN-FPI Training School on Plenoptic Sensing

Midterm Exam Solutions

The real voyage of discovery consists not in seeking new landscapes, but in having new eyes.

Geometry for Computer Vision

Homography based visual odometry with known vertical direction and weak Manhattan world assumption

Perspective Projection [2 pts]

Computer Vision I - Robust Geometry Estimation from two Cameras

A Factorization Method for Structure from Planar Motion

Camera Parameters, Calibration and Radiometry. Readings Forsyth & Ponce- Chap 1 & 2 Chap & 3.2 Chap 4

C280, Computer Vision

Epipolar Geometry and Stereo Vision

Stereo and Epipolar geometry

Transcription:

Mysteries of Parameterizing Camera Motion - Part 1 Instructor - Simon Lucey 16-623 - Advanced Computer Vision Apps

Today Motivation SO(3) Convex? Exponential Maps SL(3) Group.

Adapted from: Computer vision: models, learning and inference. Simon J.D. Prince Motivation

Motivation Taken from: H. Liu et al. Robust Keyframe-based Monocular SLAM for Augmented Reality, ISMAR 2016.

Motivation The cathedral dataset: 480 camera matrices [ i, i ] Total dof = 480 (3 + 3) = 2880 91178 3D points. Total dof = 91178 3 = 273543 Adapted from: Optimization Methods in Computer Vision. Anders Eriksson

Structure from Motion FX NX x f n (w n ; f ) 2 2 f=1 n=1 x 2D projection w 3D point extrinsics N no. of points F projection function no. of frames

Reminder: Cheat Sheet Description 3D Point 2D Point Rotation matrix Intrinsics matrix Hartley & Zisserman X x R K Prince w x Homography matrix H translation vector t

Reminder: Extrinsics w w 0 u 0 camera coordinate frame o 0 o world coordinate frame u apple apple apple apple u 0!1! 2 u x w + w 0 = Rotation Matrix Translation Vector! 3! 4 z

Reminder: Extrinsics Position w=(u,v,w) T of point in the world is generally not expressed in the frame of reference of the camera. Transform using 3D transformation or Point in frame of reference of camera Point in frame of reference of world

Structure from Motion FX NX arg min W, x f n (w n ; f ) 2 2 f=1 n=1 apple w 1 (w; ) =A( )

Weak Perspective Weak perspective is an orthographic projection plus scaling. Approximates perspective when, we assume all points on a 3D object are roughly the same distance from the camera. the size of the model in depth is small compared to the depth of the model centroid d c >> d m Computations much simpler than full perspective. d c depth of model centroid from camera d m depth of model

Structure from Motion FX NX arg min W, x f n (w n ; f ) 2 2 f=1 n=1 apple w 1 (w; ) =A( ) A( ) =s apple 1 0 0 0 1 0 [, ]

X A W 3 N 2F N Is this problem convex? Known Unknown Tomasi & Kanade. 1992 13

Today Motivation SO(3) Convex? Exponential Maps SL(3) Group.

Convex Optimization A convex function x 0 and x 1 in S D f : S D! R is one that satisfies, for any f([1 ]x 0 + x 1 ) apple [1 ]f(x 0 )+ f(x 1 ) s.t. 0 apple apple 1 f(x 0 ) f(x 1 ) x 0 x 1

Convex Objectives Common convex objective functions in computer vision, f(x) = x 2 2! L 2 penalty f(x) = x 1! L 1 penalty (Quadratic Programming Problem) (Linear Programming Problem) L 2 f(x) L 1 x

Quadratic Programming Most widely used in vision and learning. arg min x x T Px + q T x + r s.t. Gx apple h Ax = b x 2 S D Examples - SfM, Support Vector Machines, Alignment, etc.

Properties Two important properties:- 1. A convex function has a single minimum on a convex domain. S S

Properties Two important properties:- 1. A convex function has a single minimum on a convex domain. S S S is not convex!!!

Example - SO(3) is not a Convex Domain As is a rotation matrix it is constrained by the following, T = I det( ) =1 We refer to these matrices as belonging to the Special Orthogonal Group - SO(3). 1, 2 2 SO(3) 1 +(1 ) 2 2/ SO(3), 8 s.t. 0 apple apple 1

Review - Something to try In MATLAB type, >> R1 = orth(randn(3,3)); >> R1(:,end) = det(r1)*r1(:,end); >> R2 = orth(randn(3,3)); >> R2(:,end) = det(r2)*r2(:,end); If you form a new matrix as a linear combination of R1 & R2, >> R3 = 0.5*R1 + 0.5*R2; Does R3 lie in SO(3)?

Today Motivation SO(3) Convex? Exponential Maps SL(3) Group.

Review: Pinhole Camera Real camera image is inverted Instead model impossible but more convenient virtual image Adapted from: Computer vision: models, learning and inference. Simon J.D. Prince

Structure from Motion FX NX arg min W, x f n (w n ; f ) 2 2 f=1 n=1 apple w 1 (w; ) =A( )

Gauss-Newton Algorithm Gauss-Newton algorithm common strategy for optimizing non-linear least-squares problems. arg min x y F(x) 2 2 s.t. F : R N! R M Step 1: arg min x y F(x) @F(x) @x T x 2 2 Carl Friedrich Gauss Step 2: x x + x keep applying steps until x converges. Isaac Newton 26

Initialization Initialization has to be suitably close to the ground-truth for method to work. Kind of like a black hole s event horizon. You got to be inside it to be sucked in!

Gauss-Newton Optimization Can we approximate the true objective with a convex one?

Reminder: Convex Set 29

Reminder: Non-Convex Set 30

Euler Angles We could consider Euler Angles right???? R( x, y, z )=R( x )R( y )R( z ) Euler angles do form a convex set.. 0 apple apple 2 Do Euler Angles form a unique rotation? 31

Gimbal Lock 32