Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Lecture1

  • Be the first to comment

  • Be the first to like this

Lecture1

  1. 1. Image ProcessingLecture 1Introduction and Application-Gaurav Gupta-Shobhit Niranjan
  2. 2. Instructors Gaurav GuptaCan catch at : B109/Hall-1mail at: gauravg@iitk.ac.ingtalk: gauravg.84@GMAILmore information at: http://home.iitk.ac.in/~gauravg Shobhit NiranjanCan catch at : B211/Hall-1mail at: nshobhit@iitk.ac.ingtalk: shobhitn@GMAILmore information at: http://home.iitk.ac.in/~nshobhit
  3. 3. Course Structure1. Introduction to Image Processing, Application andProspects (Today)2. Introduction, Image formation, camera models andperspective geometry3. Fourier Transform theory , Convolution andCorrelation4. Color, Image enhancement Techniques5. Binary images: thresholding, moments, topologyNote: Some topics may not be in order to maintain coherency and runningtime requirements. (Bare with us …trying to teach first time !!)
  4. 4. Today Image Formation Range Transformations Point Processing Reading for this week: Gonzalez & Woods, Ch. 3
  5. 5. Image Formationf(x,y) = reflectance(x,y) * illumination(x,y)Reflectance in [0,1], illumination in [0,inf]
  6. 6. Sampling and Quantization
  7. 7. Sampling and Quantization
  8. 8. What is an image? We can think of an image as a function, f, from R2to R: f( x, y ) gives the intensity at position ( x, y ) Realistically, we expect the image only to be defined over arectangle, with a finite range: f: [a,b]x[c,d]  [0,1] A color image is just three functions pasted together.We can write this as a “vector-valued” function:( , )( , ) ( , )( , )r x yf x y g x yb x y  =   
  9. 9. Images as functions
  10. 10. What is a digital image? We usually operate on digital (discrete) images: Sample the 2D space on a regular grid Quantize each sample (round to nearest integer) If our samples are ∆ apart, we can write this as:f[i ,j] = Quantize{ f(i ∆, j ∆) } The image can now be represented as a matrix of integer values
  11. 11. Image processing An image processing operation typically defines anew image g in terms of an existing image f. We can transform either the range of f. Or the domain of f: What kinds of operations can each perform?
  12. 12. Negative
  13. 13. Log
  14. 14. Image Enhancement
  15. 15. Contrast Streching
  16. 16. Image Histograms
  17. 17. Histogram Equalization
  18. 18. Neighborhood Processing (filtering) Q: What happens if I reshuffle all pixels withinthe image? A: It’s histogram won’t change. No pointprocessing will be affected… Need spatial information to capture this.
  19. 19. Programming Assignment #1 Easy stuff to get you started withMatlab Shobhit will hold your first tutorial Topics will be from next 2 lectures
  20. 20. Applications&Research Topics
  21. 21. Document Handling
  22. 22. Signature Verification
  23. 23. Biometrics
  24. 24. Fingerprint Verification / Identification
  25. 25. Fingerprint Identification Research atUNRMinutiae MatchingDelaunay Triangulation
  26. 26. Object Recognition
  27. 27. Object Recognition Researchreference view 1 reference view 2novel view recognized
  28. 28. Indexing into Databases Shape content
  29. 29. Indexing into Databases (cont’d) Color, texture
  30. 30. Target Recognition Department of Defense (Army, Airforce,Navy)
  31. 31. Interpretation of aerial photography is a problem domain in bothcomputer vision and registration.Interpretation of Aerial Photography
  32. 32. Autonomous Vehicles Land, Underwater, Space
  33. 33. Traffic Monitoring
  34. 34. Face Detection
  35. 35. Face Recognition
  36. 36. Face Detection/Recognition Research atUNR
  37. 37. Facial Expression Recognition
  38. 38. Face Tracking
  39. 39. Face Tracking (cont’d)
  40. 40. Hand Gesture Recognition Smart Human-Computer User Interfaces Sign Language Recognition
  41. 41. Human Activity Recognition
  42. 42. Medical Applications skin cancer breast cancer
  43. 43. Morphing
  44. 44. Inserting Artificial Objects into a Scene
  45. 45. Companies In this Field In India and cameto IITK* Sarnoff Corporation Kritikal Solutions National Instruments GE Laboratories Ittiam, Bangalore Interra Systems, Noida Yahoo India (Multimedia Searching) nVidia Graphics, Pune (have high requirements) ADE Bangalore, DRDO
  46. 46. Links for Self Study and a little Play http://undergraduate.csse.uwa.edu.au/units/233.4 http://www.netnam.vn/unescocourse/computervis Book: Digital Image Processing, 2nd Editionby Gonzalez and Woods, Prentice Hall
  47. 47. Good Luck !!!

    Be the first to comment

    Login to see the comments

Views

Total views

428

On Slideshare

0

From embeds

0

Number of embeds

2

Actions

Downloads

0

Shares

0

Comments

0

Likes

0

×