Presentation is loading. Please wait.

Presentation is loading. Please wait.

Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Stanford CS223B Computer Vision, Winter 2007 Lecture 1 Course Overview Image Formation.

Similar presentations


Presentation on theme: "Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Stanford CS223B Computer Vision, Winter 2007 Lecture 1 Course Overview Image Formation."— Presentation transcript:

1

2 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Stanford CS223B Computer Vision, Winter 2007 Lecture 1 Course Overview Image Formation Professors Sebastian Thrun and Jana Kosecka CAs: Vaibhav Vaish and David Stavens

3 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Today’s Goals Learn about CS223b Get Excited about Computer Vision Learn about Image Formation (Part 1)

4 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Administrativa Time and Location Mon/Wed 11:00-12:15, Gates B1 On announcement: Fri 3:15-4:00 SCPD Televised Web site http://cs223b.stanford.edu

5 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 People Involved You: 56 students signed up Us: –Sebastian Thrun thrun@stanford.edu Office hours Thu 3-4 with appointment, Gates 154 –Jana Kosecka kosecka@cs.gmu.edu Office hours: Mon, Wed 1-2pm, Gates 258? –CA: Vaibhav Vaish Mon 2-3:30pm Gates 360 –CA: David Stavens Odd hours, Gates 254

6 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Goals To familiarize you with basic the techniques and jargon in the field To enable you to solve computer vision problems To let you experience (and appreciate!) the difficulties of real-world computer vision To get you excited!

7 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Course Requirements + Criteria You have to –Turn in all assignments (30% of final grade) –Pass the midterm (30%) –Carry out research project (40%) Late policy –Six late days (exception: midterm, project report) Teaming: –Assignments/project: up to three students –Teams may change (but watch your late days)

8 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Course Overview Basics –Image Formation and Camera Calibration –Image Features –Calibration 3D Reconstruction –Stereo –Image Mosaics, Stiching Motion –Optical Flow –Structure From Motion –Tracking Object detection and recognition –Grouping –Detection –Segmentation –Classification

9 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Course Overview

10 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Homework 1 is Online Due Jan 24, 11:59pm PST, per Email to cs223b@gmail.com

11 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 The Texts

12 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Project Deadlines Check Web site for proposals, or develop your own Team up! Dates –Feb 14: Project proposals due –Mar 14: Final report due –Mar 18: Mini-workshop on projects

13 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Projects Page

14 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 To define your own project… Find a mentor (e.g., one of the instructors) Generate project description for the Class Web site Gather data, process data Write suitable project proposal

15 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Define your own project! Learn to find cool videos on youtube.com Match images of same location at flickr.com Fly autonomous helicopter with camera

16 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Today’s Goals Learn about CS223b Get Excited about Computer Vision Learn about image formation (Part 1)

17 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Computer Graphics Image Output Model Synthetic Camera (slides courtesy of Michael Cohen)

18 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Real Scene Computer Vision Real Cameras Model Output (slides courtesy of Michael Cohen)

19 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Combined Model Real Scene Real Cameras Image Output Synthetic Camera (slides courtesy of Michael Cohen)

20 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 1:Stereo See http://schwehr.org/photoRealVR/example.html

21 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 2: Structure From Motion http://www.cs.unc.edu/Research/urbanscape

22 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 2: Structure From Motion http://www.cs.unc.edu/Research/urbanscape

23 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 2: Structure From Motion http://www.cs.unc.edu/Research/urbanscape

24 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 2: Structure From Motion http://www.cs.unc.edu/Research/urbanscape

25 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 2: Structure From Motion http://www.cs.unc.edu/Research/urbanscape

26 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 3: 3D Modeling http://www.photogrammetry.ethz.ch/research/cause/3dreconstruction3.html

27 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 4: 3D Modeling Drago Anguelov

28 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 4: 3D Modeling

29 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 4: 3D Modeling

30 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 5: Segmentation http://elib.cs.berkeley.edu/photos/classify/

31 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 6: Classification

32 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 6: Classification

33 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 7: Tracking http://www.seeingmachines.com/facelab.htm

34 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 8: Detection and Tracking David Stavens, Andrew Lookingbill, David Lieb, CS223b Winter 2004

35 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 9: Human Vision

36 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Example 9: Human Vision

37 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Excited Yet?

38 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Today’s Goals Learn about CS223b Get Excited about Computer Vision Learn about image formation (Part 1)

39 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Topics Pinhole Camera Orthographic Projection Perspective Camera Model Weak-Perspective Camera Model

40 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Pinhole Camera *many slides in this lecture from Marc Pollefeys comp256, Lect 2 -- Brunelleschi, XVth Century

41 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Perspective Projection A “similar triangle’s” approach to vision. Marc Pollefeys

42 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Implications For Perception* * A Cartoon Epistemology: http://cns-alumni.bu.edu/~slehar/cartoonepist/cartoonepist.html Same size things get smaller, we hardly notice… Parallel lines meet at a point…

43 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Perspective Projection x fZ X O -x f

44 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Consequences: Parallel lines meet There exist vanishing points Marc Pollefeys

45 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 The Effect of Perspective

46 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Vanishing points VPL VPR H VP 1 VP 2 VP 3 Different directions correspond to different vanishing points Marc Pollefeys

47 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Question How many vanishing points are there in an image? 1236100 ∞

48 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Perspective Projection fZ X O -x

49 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Weak Perspective Projection f Z O -x Z Z

50 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Generalization of Orthographic Projection When the camera is at a (roughly constant) distance from the scene, take m=1. Marc Pollefeys

51 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Pictorial Comparison  Weak perspective Perspective Marc Pollefeys

52 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Summary: Perspective Laws 1.Perspective 2.Weak perspective 3.Orthographic

53 Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Limits for pinhole cameras


Download ppt "Sebastian Thrun & Jana Kosecka CS223B Computer Vision, Winter 2007 Stanford CS223B Computer Vision, Winter 2007 Lecture 1 Course Overview Image Formation."

Similar presentations


Ads by Google