EE 5630 Advanced Digital Image Processing

Course Syllabus for Spring 2013

Abstract This course will introduce the student to practical aspects of advanced digital image processing, using specic applications to demonstrate these principles. The use of modern engineering tools such as MATLAB R to solve image processing problems is emphasized. Other solution approaches (C, C++, Java, FORTRAN, IDL, etc.) will be discussed.

Contents
1 Instructor 2 Course Goals and Objectives 3 Course Prerequisites 4 Course Grading 4.1 Breakdown of Graded Events . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Turn-In Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Grading Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Course Materials and 5.1 Textbook . . . . . 5.2 Web Sites . . . . . 5.3 Software and Other Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Resources . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 2 3 3 3 3 4 4 4 4 4 4 5 5 5 6 8

6 Miscellaneous Course Policies 6.1 Collaboration Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Academic Honesty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Disability Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Course Overview 8 Course Topic Outline References

Instructor

Dr. Cameron H. G. Wright, P.E. Oce: EN 5052 Phone: 766-6104 E-mail: chgw@uwyo.edu Oce hours: Mon and Wed: 3:304:30 PM Tue and Thu: 10:0011:30 AM

Course Goals and Objectives

Goal: The goal of this course is for the students to understand digital image processing beyond just the fundamental or introductory level, to choose appropriate image processing algorithms to achieve a desired result, to properly implement such algorithms using modern computing tools such as MATLAB, and to correctly interpret and present the results. Objectives: Students will be able to Explain and demonstrate the limitations and tradeos of various of digital image representations such as computed tomography, grayscale versus color, and tools such as wavelet transforms and image compression techniques. Display the ability to manipulate both binary and grayscale digital images using morphological lters and operators to achieve a desired result. Show how higher-level image concepts such as edge detection, segmentation, representation, and object recognition can be implemented and used. Demonstrate the programming skills to correctly implement the chosen algorithm in MATLAB, critically interpret the results, and write up a professional-quality report. Demonstrate the ability to communicate technical topics clearly, concisely, and with proper grammar.

Course Prerequisites

Satisfactory completion of EE 4530 Introduction to Digital Image Processing or an equivalent introductory image processing course is highly recommended. The student is expected to be comfortable with foundational topics such as basic linear algebra, Fourier transforms, linear signals and systems (i.e., convolution, correlation, etc.), the sampling theorem, quantization, probability (of a single variable), basic spatial domain image processing, basic frequency domain image processing, and programming in MATLAB. If the student feels rusty in any of these areas, it would be wise to review such material on your own early in the semester.

4
4.1

Course Grading
Breakdown of Graded Events

Graded material is given the following weights. Note the Final Exam is scheduled for Friday, May 10, 2013 from 8:00 AM10:00 AM. Weight Exam 1 20% Final Exam 30% Computer Projects 40% Quizzes, Homework, and/or In-Class Exercises 10% Total 100%

4.2

Turn-In Policy

All computer project assignments are due NLT 1600L (for civilians, that means no later than 4:00 PM local time ) on their respective due dates. The format for project reports is given on the course web page (in the Admin section). A note on software implementation for the computer projects: The student is not allowed to use the MATLAB Image Processing Toolbox or any other pre-written image processing (IP) software unless specically authorized by the instructor. Many, many past students would agree that writing the code yourself is the best way to understand the IP algorithms. The normal turn-in method for a computer project is electronic (as e-mail attachments) and should include both your report le (in PDF format only) and any MATLAB mles of original software you created for the project.1 Note that your reports Appendix must include as a minimum the program listings of all your original m-les used for the project, so the report can be a stand-alone document. Pay attention to the required naming conventions for your les, as given in the project assignment instructions. Late turn-ins will not be accepted unless an extraordinary situation exists. Unforeseen computer, e-mail, or printer problems do not constitute an extraordinary situationso plan ahead! Computer projects and homework almost always take longer than you expect. . .

4.3

Grading Scale

The course will be graded on the following scale, where the students overall average in the course, as a percentage, is represented by x. No grade curving will be used. Grade Average (%) A 90 x 100 B 80 x < 90 C 70 x < 80 D 60 x < 70 F 0 x < 60
Be sure to include in the comments any needed explanation about your program, its operation, or potential problems running it.
1

5
5.1

Course Materials and Resources

Textbook

The primary textbook used for this course, by R. C. Gonzalez and R. E. Woods, is listed as [1] in the reference section of this syllabus. This is the same book used for the prerequisite EE 4530 and it is by far the most popular image processing text in the world. You should use the book (and its companion website) extensively in this course. For some topics, handouts may also be used to supplement the text.

5.2

Web Sites

For this course, we will make frequent use of the course web site located at http://wwweng.uwyo.edu/classes/sp2013/ee5630/ You can also navigate to this site from the CEAS main pageCurrent StudentsWebbed Classes. A copy of this syllabus, assignments, data les, and other supplemental material will be posted to this web site. You are responsible for checking it regularly. Another web site you may nd very helpful is the textbook companion web site: http://www.imageprocessingplace.com/. This is a richly populated web site with all the gures in the book, tutorials, links to many other sites, etc. Take advantage of it!

5.3

Software and Other Resources

Each student must become familiar with the following software: MATLAB, version 7 (or higher), The MathWorks, Inc. This software is available in the ESIG Labs (EN 1034, 1039 and 1041) and in the ECE Department Student Computing Lab (EN 5038). Note that there are signicant price reductions on MATLAB and its Toolboxes for student purchasers if you wish to work at home.

6
6.1

Miscellaneous Course Policies

Collaboration Policy

For all computer projects, you will work as a team of two students.2 You may also consult with other students currently enrolled in this course who are not on your team, but you must document any such help you receive (e.g., as comments directly in the source code of
If there is an odd number of students in the class, one and only one team will have just a single member. This team approach is a reection of the team-oriented nature of essentially every engineering (and computer science) type of job in todays workforce.
2

your program and in the References section of your project report). No comments mean you are submitting the project as totally your own teams work; my assumption will always be that you are an honorable person unless you cause me to believe otherwise. Simply copying another teams assignment is not allowedthe actual item you turn in must ultimately be your own teams work. You may be called upon to explain to me in person and in detail how certain answers were obtained or how your software works. Exams and quizzes must always be the students own work.

6.2

Academic Honesty

The University of Wyoming is built upon a strong foundation of integrity, respect and trust. All members of the University community have a responsibility to be honest and have the right to expect honesty from others. Any form of academic dishonesty is unacceptable to our community and will not be tolerated. [excerpted from the UW General Bulletin] All persons should report suspected violations of standards of academic honesty to the instructor, department head, or dean. See UW Regulation 6-802, Procedures and Authorized University Actions in Cases of Student Academic Dishonesty. You can read this and all other University regulations at: http://www.uwyo.edu/generalcounsel/index.html

6.3

Disability Statement

If you have a physical, learning, or psychological disability and require accommodations, please let the instructor know as soon as possible. You must register with, and provide documentation of your disability, to University Disability Support Services (UDSS) in SEO, room 330 Knight Hall.

Course Overview

Welcome to Advanced Digital Image Processing! This course will continue where EE 4530 left o, and prepare you to deal competently with a wide range image processing challenges that may confront you in the future. The subject had consistently been described as fun and enjoyable by most students, but it isnt just playing with pictures! There is a considerable need for quantitative, mathematical descriptions and manipulations of imagesso be prepared for that. However, the emphasis in this course is on the underlying principles and practical applications, rather than esoteric theory or arcane derivations. Just like with anything worthwhile in life, if you arent willing to put in the time, you wont ever get good at it. Be prepared to devote considerable time to this class. I promise to be sensitive to the time requirements of every assignment I give you, but you have to put forth the eort. As Robert E. Heinlein was known to say, TANSTAFL. Some recommendations for success in this class which you might want to consider. . . Dont miss class. New material is covered each lecture. Attendance isnt strictly required, but if you miss class, you are responsible for covering the missed material on your own. Repeat lectures will not be given during oce hours.

Read in advance. Based upon the topics were covering or are about to cover, use the table of contents and/or index of your text (and any supplementary references you use) to read about these topics in advance. Start your projects early. Give yourself some time to consider the problem and determine whether or not you need instructor assistance. Last-minute questions are a bad idea. Do your best on the projects. They comprise 40% of your grade! Ask questions. This includes during class, during discussions, and during oce hours. I dont like a silent classfeel free to ask questions or make reasonable comments at will (but no distracting side conversations). Dont arrive late for class. If you know youll be delayed (or absent) for some reason, just let me know ahead of time in person or via e-mail. Its the courteous and adult thing to do. Lets have fun in this class! I like lively, attentive, alert students with an active sense of humor. Its more fun that way for all of us. . .

Course Topic Outline

The following outline lists the basic sequence of course topics. We will begin with a quick review of topics from EE 4530, then concentrate on the last part of Chapter 5 (reconstruction from projections, as in computerized tomographic imaging used for CT scans, PET scans, MRIs, etc.). Following this, well proceed to Chapters 612 in the text. The speed and depth of coverage will depend to some degree upon the students, but remember: this is a graduate-level class. That means a greater amount of independent reading and more work outside of class is expected of you than for an undergraduate class. More specic lesson-by-lesson guidance will be given when needed as the semester progresses. Review of basic digital image processing (Chap. 15) Spatial domain image processing Frequency domain image processing Computerized tomographic imaging (end of Chap 5 and handouts) History and applications of computed tomography Mathematics underlying computerized tomographic imaging Methods of forming a computed tomography image Color models, representation, and image processing (Chap. 6) RGB, CMYK, HSI models, and pseudocolor Important color models not in book: e.g., YIQ, YUV, YPbPr, YCbCr 6

 Closely related topics not in book: color models versus video/monitor standards (e.g., NTSC, PAL, SECAM, composite, S-video, component, VGA, ATSC, DVB, DVI, HDMI; interlaced versus progressive scanning; SD versus HD) Color transformations Color image ltering and enhancement Wavelets and multiresolution processing (Chap. 7) Multiresolution representations Subband coding and lter banks Not in book: bandpass sampling (related to subband coding) Wavelet transforms in 1-D and 2-D Image compression (Chap. 8) Information content (entropy) of an image Lossless versus lossy compression algorithms Compression standards (including the new JPEG XR, not in the book) Not in book: compressive sampling (a very hot topic today!) Morphological image processing (Chap. 9) Intro to image morphology Binary operations Grayscale operations Applications of morphological image processing Image segmentation (Chap. 10) Edge detection Edge linking and boundary detection Thresholding Other methods of segmentation Image representation and object recognition (Chap. 1112) Descriptors: boundaries, regions, global Pattern recognition as applied to images

References
[1] R. C. Gonzalez and R. E. Woods, Digital Image Processing, Prentice Hall, 3rd ed., 2008. Annotation: This is the number one image processing text in the world, with an amazingly complete companion website. [2] J. C. Russ, The Image Processing Handbook. CRC Press, 5th ed., 2006. Annotation: A fairly useful book, long on breadth but short on depth. [3] R. C. Gonzalez, R. E. Woods, and S. L. Eddins, Digital Image Processing Using MATLAB, Prentice Hall, 2004. Annotation: Written by the authors of your textbook and a person from The Mathworks (creators of Matlab), this book is much lighter on theory but heavy on specic Matlab examples. There is an excellent companion web site. [4] A. Rosenfeld and A. C. Kak, Digital Picture Processing, vol. 1 and 2. Academic Press, 2nd ed., 1982. Annotation: While a bit dated in some respects and hard to nd, these volumes are still considered by many to be required books to have in your personal IP library. An excellent treatment of computed tomography in Volume 1. [5] A. C. Kak and M. Slaney, Principles of Computerized Tomographic Imaging. SIAM Press, 2001. Annotation: A somewhat newer and expanded treatment of the material found in Chapter 8 (volume 1) of the Rosenfeld and Kak classic, Digital Picture Processing. [6] M. Sonka, V. Hlavac, and R. Boyle, Image Processing, Analysis, and Machine Vision. Thomson Learning, 3rd ed., 2008. Annotation: A mathematically rigorous IP book that skims over some of the basics to make room for some very interesting advanced topics. Good for graduate students. [7] K. R. Castleman, Digital Image Processing. Prentice Hall, 1979. Annotation: A good introductory text, similar in scope to your textbook. [8] W. K. Pratt, Digital Image Processing. John Wiley & Sons, 3rd ed., 2001. Annotation: Another comprehensive introductory IP book, similar in scope to your textbook. The ISO standard PIKS software for IP (written in C) and an image database is included on a CD-ROM. [9] D. A. Forsyth and J. Ponce, Computer Vision: A Modern Approach. Prentice Hall, 2003. Annotation: Contains introductory material on both lower-level image processing and higher-level computer vision. [10] R. M. Haralick and L. G. Shapiro, Computer and Robot Vision, vol. 1 and 2. AddisonWesley, 1992. Annotation: A very complete two volume set, but a bit hard to read as an introduction. After the divorce, Shapiros later book with Stockman is less complete but more approachable. [11] L. G. Shapiro and G. C. Stockman, Computer Vision. Prentice Hall, 2001. Annotation: More advanced than your textbook, but covers some good foundation for object recognition. 8

[12] E. R. Davies, Machine Vision: Theory, Algorithms, Practicalities. Morgan Kaufmann (Elsevier), 3rd ed., 2005. Annotation: A huge (nearly 1,000 pages!), comprehensive book by the well-known British author. [13] S. E. Umbaugh, Computer Imaging: Digital Image Analysis and Processing. CRC Press, 2005. Annotation: An introductory text that is light on the math but heavy on examples, the book includes the ANSI-C library and GUI skeleton software called CVIPtools on a CD-ROM. [14] D. Marr, Vision, W. H. Freeman and Company, 1982. Annotation: While Marr died soon after this book was published, it is still revered as one of the best references on the entire phenomonon of vision and imaging. [15] E. R. Dougherty and R. A. Lotufo, Hands-On Morphological Image Processing. SPIE Press, 2003. Annotation: The best tutorial on morphological IP Ive found so far! [16] J. Nakamura, Image Sensors and Signal Processing for Digital Still Cameras, CRC Press, 2006. [17] R. H. Vollmerhausen, D. A. Reago, Jr., and R. G. Driggers, Analysis and Evaluation of Sampled Imaging Systems, SPIE Press, 2010. Annotation: updated excellent treatment of aliasing and other image degradations [18] K. R. Castleman, Resolution and sampling requirements for digital image processing, analysis, and display, in Shotton [19], ch. 3, pp. 7194. [19] D. Shotton, ed., Electronic Light Microscopy: Techniques in Modern Biomedical Microscopy. Wiley-Liss, 1993. [20] P. J. Van Fleet, Discrete Wavelet Transformations. John Wiley & Sons, 2008. Annotation: A highly readable introduction to wavelet theory and applications. [21] G. Strang and T. Nguyen, Wavelets and Filter Banks. Wellesley-Cambridge Press, 1997. Annotation: Well written text that links wavelets to more well-known DSP lter theory. [22] M. Vetterli and J. Kovaevi, Wavelets and Subband Coding. Prentice Hall, 1995. c c Annotation: An often cited text that is mathematically rigorous and thorough, albeit not the easiest book to read on the subject. [23] M. Rabbani and P. W. Jones, Digital Image Compression Techniques. SPIE Press, 1991. Annotation: A rigorous treatment of this very mathematical subject, but surprisingly readable. [24] G. C. Holst, Electro-Optical Imaging System Performance, SPIE Press, 3rd ed., 2003. Annotation: A non-rigorous but practical treatment of imaging system performance, going beyond just the sensors. A very readable discussion of radiometry.

[25] G. C. Holst, Sampling, Aliasing, and Data Fidelity for Electronic Imaging Systems, Communications, and Data Acquisition, SPIE Press, 1998. Annotation: Not particularly rigorous mathematically, but it provides a good seat of the pants intuitive approach to the subject. [26] G. C. Holst, CCD Arrays, Cameras, and Displays, SPIE Press, 1996. Annotation: Similar in some respects to the authors 2003 text, this book discusses mainly CCD sensors and cameras. [27] K. Aizawa, K. Sakaue, and Y. Suenaga, eds., Image Processing Technologies: Algorithms, Sensors, and Applications. Marcel Dekker, 2004. [28] K. B. Benson and J. Whitaker, Television Engineering Handbook. McGraw-Hill, revised ed., 1992. Annotation: An excellent resource with regard to video standards and related technology. [29] T. Fukinuki, Television: Past, present, and future, Proceedings of the IEEE, vol. 86, pp. 9981004, May 1998. Annotation: A very readable combinaton of TV history and the HDTV future. [30] S. Inou and K. R. Spring, Video Microscopy: The Fundamentals. Plenum Press, e 2nd ed., 1997. Annotation: It may not be evident from the title, but this book contains one of the best treatments of video Ive ever come across. [31] T. S. Rzeszewski, ed., Digital Video. IEEE Press, 1995. Annotation: A good collection of seminal papers on all types of video. [32] A. F. Inglis and A. C. Luther, Video Engineering. McGraw-Hill, 2nd ed., 1996. Annotation: An excellent video reference, fewer topics but more rigor than Bensons handbook. [33] S. F. Barrett, C. H. G. Wright, H. Zwick, M. Wilcox, B. A. Rockwell, and E. Naess, Eciently tracking a moving object in two-dimensional image space, SPIE Journal of Electronic Imaging, vol. 10, pp. 785793, July 2001. [34] C. H. G. Wright, S. F. Barrett, D. J. Pack, T. R. Schei, J. R. Anderson, and M. J. Wilcox, Computational image processing for a computer vision system using biomimetic sensors and eigenspace object models, in Proceedings of the SPIE 16th International Symposium on Electronic Imaging (SPIE 5299), pp. 327336, Jan. 2004. [35] R. S. Prabhakara, C. H. G. Wright, and S. F. Barrett, Motion detection: a biomimetic vision sensor versus a CCD camera sensor, IEEE Sensors J., vol. 12, pp. 298307, Feb. 2012. (Invited paper).

. . . one more page to the end . . .

10

Finally. . .
Congratulations on reading this far! Lesser mortals gave up several pages earlier. As a reward, perhaps these quotes will inspire you in a positive manner. . .

To study, and when the occasion arises to put what one has learned into practiceis that not deeply satisfying? Confucius

The purpose of computing is insight, not numbers. Richard W. Hamming

Great minds discuss ideas, average minds discuss events, small minds discuss people. Admiral Hyman G. Rickover (father of the nuclear Navy)

If I have seen further than others, it is by standing on the shoulders of giants. Sir Isaac Newton

You do not really understand something unless you can explain it to your grandmother. Albert Einstein

That which does not kill us makes us stronger. Friedrich Nietzsche

11