CMSC 426
Image Processing (Computer Vision)
David Jacobs
Spring 2004

 

   Announcement

 
  
 
General Information


 

Class Time 
Tue, Thu 11:00-12:15
Room CSI 1122
Course Info Syllabus
Text Introductory Techniques for 3-D Computer Vision by Trucco and Verri. 
Personnel

  Instructor Teaching Assistant
Name David Jacobs Konstantinos Bitsakos
Email djacobs@cs.umd.edu kbits@cs.umd.edu
Office AVW 4421 AVW 1112
Office hours Tue 2:00-3:00, Wed 2:00-3:00  Tue 12:30-1:30, Thurs 9:30-10:30

If you cannot make these office hours, please send email to arrange another time. 

Below is a tentative schedule for problem sets and lectures.
 


  Assigned Due
  Problem Set
    Problem Set 1 : PDF
Jan. 29th Feb. 5
    Problem Set 2PDFWord

Matlab routines: ladderdisplay_image_ptsmovie_pts

Feb. 5 Feb. 12
    Problem Set 3: PDFWord Feb. 12 Feb. 19
    Problem Set 4: PDFWord

Zipped file with all matlab code

Or, individual files: extractmin.m, filterimage.m, imagetoolbox.m, insert.m, ps4.m, scissor.m, swan.jpg

This problem set is based on one developed by Steve Seitz at the University of Washington.  His description of the problem, and sample results can be found here. The problem set is based on the reading:: Intelligent Scissors for Image Composition, by Eric Mortensen and William Barrett, SIGGRAPH '95.

Feb. 19 Mar. 4
    Problem Set 5: PDF Word

This problem set is based on ``Texture Synthesis by Non-parametric Sampling'' by Efros and Leung.  See their web page for the paper, pseudocode and examples.

Here is a function checkerboard.m and a sample image, brickbw.jpg, to use to test your program.

Mar. 11 Apr. 1
    Problem Set 6: PDFWord. I1L.jpg I1R.jpg  T3bw.jpgT4bw.jpgI3.jpg Apr. 1 Apr. 15
      Problem Set 7: WordWeb page Apr. 29 May 6
A hardcopy of all solutions and code should be turned in by end of class on the assigned date.  If the assignment involves writing Matlab code, all code should be emailed to the TA also, so that it can be run, if needed.  Please tar all matlab files into one file named student_name_psX and email it to kbits@wam.umd.edu

Problems not completed on time can be turned in one day late, for 90% credit, or four days late for 70% credit.

Class slides will typically be posted here, usually minutes before class begins.  In most cases, you can find a good approximation to these slides from last year's class, accessible from my home page.
 

  Class Slides    Class  1 :      Introduction (Jan. 27): Powerpoint Web page READINGS
   Class  2 :      Image Formation (Jan. 29): PowerpointWeb page T&V Chapter 1, Sections 2.1 and 2.2 (only skim 2.2.2).
   Class  3 :      Matlab Tutorial  and Linear Algebra (Feb. 3) PowerpointWeb page Matlab
   Class  4:       Filtering (Feb. 5) PowerpointWeb PageMatlab T&V Chapter 3.
   Class  5  :     Edge Detection -1 (Feb. 10) Powerpoint Web page Matlab T&V Sections 4.1, 4.2
   Class  6 :      Edge Detection -2 (Feb. 12) 
   Class  7 :      SNAKEs (finding boundaries) (Feb. 17) PowerpointWeb page T&V Section 5.4
   Class  8:       Corners  Powerpoint Web page T&V Section 4.3
   Class  9 :      Texture Powerpoint Web page See chapter 9, Forsyth and Ponce
   Class 10 :     Texture - Advanced 
   Class 12:   Perspectives on Vision (Marr and Bayesian Reasoning) (Mar. 4)
   Class 12a :    Review for Quiz Friday, March 5, 1:30-3:00, A.V. Williams, Room 4424.
   Class 13:     QUIZ I (Mar. 9) - Covers projection, edge detection, corners, and SNAKES.  
   Class 14:     Go over quiz.  Catch up. (Mar. 11)
   Class 15:     Human perceptual grouping (March 16) Powerpoint 

Web page
                                           Reading "Laws of organization in perceptual forms" at http://psy.ed.asu.edu/~classics/Wertheimer/Forms/forms.htm

   Class 16:     Grouping: clustering, RANSAC, Hough (Mar. 18). T&V 5.1, 5.2
   Class 17 :     Grouping: E-M or Spectral Methods -  (March 30)  Powerpoint Web page .  Matlab for K-means:  kmeans.m kmean_demo.m Matlab for E-M  tutorial.m tutorial2.mplot_em
   Class 18:      Stereo (1) -  (April 1) PowerpointWeb page T&V 7.1, 7.2, 7.3.1, 7.3.2, 7.3.7, 7.4 through 7.4.1.
   Class 19  :     Stereo (2) -  (April 6) 
   Class 20 :     Structure from Motion (1 ) -  (April 8) PowerpointWeb Page tutorial.m T&V 8.5.1
   Class 21       Structure from Motion (2 ) -  (April 13) 
   Class 22  :     Structure from Motion (3) -  (April 15) 
   Class 23        Matching and Optical Flow I -- Quiz Review (Apr. 20) PowerpointWeb page T&V 8.3 and 8.4 (but not Kalman tracking).
   Class 24  :    QUIZ II covers texture, grouping, stereo, structure-from-motion (Apr. 22) Practice Quiz (word) (pdf).
   Class 25        Matching and Optical Flow I (continued) -- Quiz Answers (Apr. 27)
   Class 26        Matching and Optical Flow II    (Apr. 29)
   Class 27        Recognition I  (May 4) Powerpoint   Web page
     Class 28        Recognition II  (May 6)  
    Class 29         Summary  (May 11)
    FINAL   Thursday, May 13, 8am in the classroom.  Final Review (word) (html)

 

Useful Links

 Image Processing Learning resources:
    http://www.dai.ed.ac.uk/HIPR2/index.htm

 MATLAB resources:

  Introductory Tutorials

  Slightly more advanced Tutorials   More complete references/tutorials/FAQs  Computer Vision related sites: