Stereo Vision Egon Elbre Hans Mesalu general stuff about this 3D - - PowerPoint PPT Presentation

stereo vision
SMART_READER_LITE
LIVE PREVIEW

Stereo Vision Egon Elbre Hans Mesalu general stuff about this 3D - - PowerPoint PPT Presentation

Stereo Vision Egon Elbre Hans Mesalu general stuff about this 3D thing why? applications games movies simulators robotics product presentations architecture visualization virtual television studios virtual presence for video


slide-1
SLIDE 1

Stereo Vision

Egon Elbre Hans Mäesalu

slide-2
SLIDE 2

general stuff about this 3D thing

slide-3
SLIDE 3

why?

slide-4
SLIDE 4

applications

games movies simulators robotics product presentations architecture visualization virtual television studios virtual presence for video communications general virtual reality applications

slide-5
SLIDE 5

Single-View Geometry

slide-6
SLIDE 6
  • rthographic projection
slide-7
SLIDE 7

perspective projection

slide-8
SLIDE 8

IRL

slide-9
SLIDE 9

simple camera projection

slide-10
SLIDE 10

just think of it as magic

slide-11
SLIDE 11

extrinsic/intrinsic camera calibration matrix

slide-12
SLIDE 12

finding 3D point with triangulation

assuming we know where the cameras are

slide-13
SLIDE 13

finding a depth map

slide-14
SLIDE 14

image rectification

slide-15
SLIDE 15

finding matching points simpler rectification gives us

slide-16
SLIDE 16

step 1 - get the pictures

slide-17
SLIDE 17

step 2 - find some interesting points

slide-18
SLIDE 18

step 3 - guess similar points

slide-19
SLIDE 19

step 4 - remove outliers (RANSAC)

slide-20
SLIDE 20

RANSAC RANdom SAmpling Consensus

slide-21
SLIDE 21

rectify

slide-22
SLIDE 22

disparity map

slide-23
SLIDE 23
slide-24
SLIDE 24

demo video

slide-25
SLIDE 25

Introduction to Epipolar Geometry

slide-26
SLIDE 26

terminology

slide-27
SLIDE 27
slide-28
SLIDE 28

F(undamental)-matrix

slide-29
SLIDE 29

E(ssential)-matrix

slide-30
SLIDE 30

and finally...

slide-31
SLIDE 31

reconstruction from two views only

  • 1. identify a number (at least 8) of point correspondences
  • 2. form linear equations based on x'TFx=0 formula
  • 3. find the solution F for those equations
  • 4. compute P, P' camera matrices from F
  • 5. given to cameras P, P' and corresponding point pairs

triangulate the 3D point X we know how to do 1 and 5 we won't discuss 2, 3 as it's about solving some linear equations and no one will remember it after the lecture anyway about 4 - well that's complicated

note! can't be done uniquely due to some ambiguity

slide-32
SLIDE 32

further reading

"Multiple View Geometry in Computer Vision", Richard Hartley and Andrew Zisserman "Uncertain Projective Geometry: Statistical Reasoning for Polyhedral Object Reconstruction", Stephan Heuel "Computer Vision: Algorithms and Applications", Richard Szelinski "Learning OpenCV", Gary Bradski and Adrian Kaehler