Winter 2004
ECE 178: Introduction to Digital Image Processing

Announcements:

Some sample questions and solutions from previous years.

Exam Week Office Hours:

TA: 11am - 1PM on Thursday, 03/18.

BSM: 4-5PM, WEDNESDAY, 03/17, 4-5PM Thursday, 03/18.

EXAM: FRIDAY, 8am-11am, 03/19.

 

Project Groups--updated to include the meeting schedules

Lecture Notes  HomeWorks  Discussions Weekly Schedule Steganogrphy Project

Instructor: 

B. S. Manjunath, manj at ece. Rm 3157, Engineering I. Tel: 805.893.7112.

Lectures: Tuesdays and Thursdays, 0800-0915 hrs., WEBB 1100.

Teaching Assistants: 

Christopher Utley, cutley00 at umail.
Evan Ruzanski, ruzanski at ece.ucsb.edu.
Srivatsan Pallavram. srivathsan at ece.ucsb.edu.

Office Hours

Manjunath: Mondays 2PM-3PM and Tuesdays 930-1030AM, or by appointment (please call 805 893 7112).

TA: 

M 10-11AM (ECI Lab, TA-E)
T 4-5 PM (TA-S)
W 2-4 PM (TA-C)
R 2-3 PM (TA-S)
R 3-4 PM (ECI LAB, TA-S))
F 2-4 PM (TA-E)
F 3-4 PM (ECI LAB, TA-C)

Discussion Session: Friday 12-1250 PM and 1-150 PM. 

Text Book: Digital Image Processing by Gonzalez and Woods. The book's web site contains pointers to additional resources and you are encouraged to take a look. Also, you may want to take a look at last year's course web page for more information.

About the course: ECE 178 is an introductory course in image processing. In this course, you will learn about digital images and how you can manipulate them.  Open to students in Engineering. You should have good background in basic calculus. Preliminary topics such as basics of linear systems, linear and circular convolution, and 2-D Fourier transform will be introduced during the first two weeks. You are expected to learn and use MATLAB and the Image Processing Toolbox for your programming assignments. No prior knowledge of MATLAB is required. 

Grading Ploicy: 20% HWs, 20% Mid-term examinations, 20% course project and 40% for the final examination. First mid-term examination will be in class on Feb 3rd. All home-works are required (a non-submission will affect your grade non-linearly). The home works are due by 11:59 PM on the day they are due (typically, on a Friday). Those that are received within 48 hours of the due date will get 50% of the points. After 48 hours, they will not receive any credit.

Lecture Notes

NOTE: Print version with six slides per page are available as well. Print the "handouts".

Week 1 - Introduction
  • Lecture 1 (handout): course objectives; introduction.Lecture 2 (handout): will continue the introduction, starting with a quick overview of MATLAB for image processing. We will conclude this lecture with a brief review of sampling and quantization (section 2.4) and relationship between pixels (section 2.5). Homework #1 is due January 16. Read Chapters 1 and 2. This requires MATLAB programming, so start early! Using MATLAB for image processing. 
  • Mathworks has a nice web page on the image processing toolbox. See also their demo page on some interesting demonstrations of using the IP toolkit. In particular, check out the deblurring demos that I showed in class.

Week 2 - Linear Shift-Invariant Systems: an overview

  • READING: Review chapters . This review chapter provides some of the background material that you might find useful this quarter. In particular, I recommend reviewing chapter 3 that discusses linear systems.
  • Lecture 3 (handout) gives an overview of continuous time linear time invariant systems. 
  • Lecture 4 (handout): We will conclude this discussion with an introduction to the 2-D linear systems (discrete) and some examples.

Week 3 -  Image Sampling and Quantization

Week 4 - 5: 2D Discrete Fourier Transform and Wavelets

  • Lecture 7 (handout).  2-D Discrete Fourier Transform. Your text has a brief introduction to Fourier transform in Chapter 4. 
  • Lecture 8: DFT discussion continued.
  • Lecture 9: Midterm I
  • Lecture 10: A quick tour of wavelets.

Week 6-7: Image Enhancement Techniques

Lecture 15: Feb 24, 2004. Midterm II : Topics include DFT and Enhancement.

Week 8-10: Image Compression

Week 10: Project Presentations & Review

  • Each project group will get 10 mins to present their project (about 5 slides).

 

Homeworks and Solutions

Steganography Project

Motivation:   The internet allows for easy dissemination of information over large areas. This is both a blessing and a curse since friends all over the world can view your information but so can everyone else.   Encrypting data has been the most popular approach to protecting information but this protection can be broken with enough computational power.   An alternate approach to encrypting data would be to hide it by making this information look like something else.   This way only friends would realize its true content.   In particular, if the important data is hidden inside of an image then everyone but your friends would view it as a picture.   At the same time your friends could still retrieve the true information.   This technique is often called data hiding or stenography and you will be implementing the technique in the class project. A wealth of information on this topic can be found at http://www.jjtc.com/Steganography. You can also find some recent technical papers describing the work on this topic at UCSB at http://vision.ece.ucsb.edu.

ECE 178 Project Goals:   Form groups of no more than 5 students per group (this is your responsibility). Each group is responsible for completing the following two goals.   As the deadline approaches more details will be given out about the format of the input and outputs of the Matlab programs.

•  Implement a technique in Matlab* that hides and recovers information bits hidden in an image.   This technique should perform the hiding such that the resulting image does not look degraded compared to the original.   You should pick a technique that will minimize its susceptibility to data “attacks” which are described below. (* or any other software environment that all your group members are comfortable with.)

•  Come up with a transformation that can be performed on the images with hidden data that will preserve the image quality but destroy the hidden signal. You want this data “attack” to be severe enough to destroy other groups' approaches yet your implementation should survive.

Grading:   Your group will be graded based on the following criteria.   It is important to note that while your group's grade will be largely dependent on the quality of work done, your group is competing against the other groups in the class for the "bonus" points.

•  The quality of the image carrying data

•  The amount of data carried

•  The ability of your technique to survive the class's data attacks.

Required:

1) Schedule meeting with the instructor to discuss the project details and goals.

2) Brief presentation of your project during the final week.

3) Demonstrate your project in the ECI lab to the instructor. Time TBD.

4) A brief project report due on the last day of instruction.

 

Discussion Session Handouts

 

 

 

Flashed With FrontFX FrontFlash For Frontpage By XZAKT Media.