Lecture 1 Introduction to Computer Vision. Lin ZHANG, PhD School of Software Engineering, Tongji University Spring 2014

Transcription

1 Lecture 1 Introduction to Computer Vision Lin ZHANG, PhD School of Software Engineering, Tongji University Spring 2014

2 Course Info Contact Information Room 314, Jishi Building Tel: TA: Lida LI, QQ: Course information can be found at

3 Materials Major materials My slides References Some papers Milan Sonka, Vaclav Hlavac, and Roger Boyle, Image Processing, Analysis, and Machine Vision, Thomson, 2008 D.A. Forsyth and J. Ponce, Computer Vision: A Modern Approach, Pearson Education, Inc., 2003

4 Examination Homework 45%: 3 times, and each time 15%. Project 50%: 2 or 3 people for one group Attendance 5% (being absent >=5 times, you will fail this course) Bonus 5%: being active in class and answering my questions correctly

5 Today What is computer vision? Course overview Course requirement

6 What is computer vision? To bridge the gap between pixels and meaning Source: S. Narasimhan What we see What a computer sees

7 What is computer vision? Source: Feifei Li

8 Human vision sclera choroid blind spot

9 Human vision

10

11 What is it related to? Source: Feifei Li

12 Vision as a measurement device Real time stereo Structure from motion Reconstruction from Internet photo collections NASA Mars Rover Pollefeys et al. Goesele et al.

13 Vision as a source of semantic information slide credit: Fei Fei, Fergus & Torralba

14 Object categorization sky building flag banner bus face street lamp bus wall cars slide credit: Fei Fei, Fergus & Torralba

15 Scene and context categorization outdoor city traffic slide credit: Fei Fei, Fergus & Torralba

16 Why study computer vision? Source: Lazebnik

17 Why study computer vision? Vision is useful: Images and video are everywhere! Personal photo albums Movies, news, sports Surveillance and security Medical and scientific images

18 Why computer vision is difficult? Challenges: viewpoint variation Michelangelo

19 Why computer vision is difficult? Challenges: illumination

20 Why computer vision is difficult? Challenges: scale slide credit: Fei Fei, Fergus & Torralba

21 Why computer vision is difficult? Challenges: deformation Xu, Beihong 1943 Source: Feifei Li

22 Why computer vision is difficult? Challenges: occlusion Magritte, 1957

23 Why computer vision is difficult? Challenges: background clutter

24 Why computer vision is difficult? Challenges: Motion

25 Why computer vision is difficult? Challenges: object intra class variation Source: Feifei Li

26 Challenges or Opportunities? Images are confusing, but they also reveal the structure of the world through numerous cues Our job is to interpret the cues! Image source: J. Koenderink

27 Depth cues: Linear perspective

28 Depth cues: Aerial perspective

29 Depth ordering cues: Occlusion

30 Shape cues: Texture gradient

31 Grouping cues: Similarity (color, texture, proximity)

32 Typical CV applications

33 Earth Viewers (3D modeling) Image from Baidu 3D Map

34 Photosynth Project products of students from 2009 Media&Arts

35 Structure from motion Bundler: Structure from Motion (SfM) for Unordered Image Collections (

36 Optical character recognition (OCR) Technology to convert scanned docs to text If you have a scanner, it probably came with OCR software Digit recognition, AT&T labs License plate readers Source: S. Seitz

37 Face detection Many new digital cameras now detect faces Canon, Sony, Fuji, Source: S. Seitz

38 Smile detection Source: S. Seitz

39 Vision-based biometrics How the Afghan Girl was Identified by Her Iris Patterns

40 Login without a password Palmprint system Fingerprint scanners on many new laptops, other devices Finger Knuckle Print system

41 Face verification National Stadium, Beijing Olympic Games, 2008

42 Object recognition (in mobile phones) Source: S. Seitz

43 Special effects: motion capture Source: S. Seitz Pirates of the Carribean, Industrial Light and Magic

44 Vision in space NASA'S Mars Exploration Rover Spirit captured this westward view from atop a low plateau where Spirit spent the closing months of Vision systems (JPL) used for several tasks Panorama stitching 3D terrain modeling Obstacle detection, position tracking For more, read Computer Vision on Mars by Matthies et al.

45 Robotics NASA s Mars Spirit Rover

46 Robotics Itti s system

47 Medical imaging 3D imaging MRI, CT Image guided surgery Grimson et al., MIT

48 Course content (just a plan) Introduction Image filtering Local interest point detectors Local feature descriptors and matching Camera models Biometrics: Theories and applications Face detection and face recognition Texture Object recognition: BoW model 3D shape recognition* Background subtraction*

49 Some tips Prerequisites Linear algebra Calculus Matlab Programming C++ Programming Knowledge sources IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) IEEE Transactions on Image Processing (TIP) International Journal of Computer Vision (IJCV) IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) IEEE International Conference on Computer Vision (ICCV) European Conference on Computer Vision (ECCV)

50 Thanks for your attention