CS 583: Introduction to Computer Vision
Fall 2009
[Announcements] [Syllabus] [Lectures] [Assignments] [Resources]
Time/Room
Monday 6:00-8:50PM@Crossings 149
Instructor
Ko Nishino
e-mail: kon drexel.edu
office: University Crossing 108
phone: (215) 895-2678
office hours: Tuesday 2:00-3:00 or by e-mail appointment
TA
Geoffrey Oxholm
e-mail:geoff.oxholmdrexel.edu
office: University Crossing 147
office hours: Thursday 2:00-3:00 or by e-mail appointment
Announcements
[09/21/2009] No class next week (9/28) (instructor out of town to
attend a conference)
[09/21/2009] First class will be 09/21 (Mon)
Syllabus
Overview
The goal of computer vision is to enable computers see the world.
By using a camera as the eye of a computer, studies in computer
vision seek to develop better means to capture and extract useful
visual information from images and videos and to use such
information to automatically interpret the beautiful world
surrounding us. This course provides an introduction to computer
vision. The first half of this course will focus on fundamental
models and algorithms in computer vision, including such topics as
image formation, image sensing, image filtering, edge extraction,
brightness and reflectance. In the second half, we will mainly
focus on computer vision applications, including various algorithms
for reconstructing 3D shape (shape-from-X, stereo, photometric
stereo), and recognizing objects in images.
Objectives
This course aims for students to (1) understand and apply
fundamental mathematical and computational techniques in computer
vision and (2) implement basic computer vision applications.
Prerequisites
There is no official prerequisite. However, basic
(undergraduate-level) understanding of Linear Algebra and Calculus
will be necessary. For the assignments, one will need to program in
C/C++ (example skeleton codes will be prepared in C++). Assignments
will require access to a digital camera. Students are encouraged to
purchase one (an inexpensive one will suffice) if he/she does not
own one. A few cameras can be checked out from the lab, too.
Topics
The following is a list of topics that will be covered in this
course. The timeline is preliminary and will most likely change.
|
Week 1
|
9/21
|
Introduction, Image Formation, Image Sensing
|
|
Week 2
|
9/28
|
No class (instructor attending a conference)
|
|
Week 3
|
10/5
|
Camera Models, Projective Geometry
|
|
Week 4
|
10/12
|
University Holiday (Columbus Day)
|
|
Week 5
|
10/19
|
Image Filtering
|
|
Week 6
|
10/26
|
Edge Detection, Motion
|
|
Week 7
|
11/2
|
Midterm Exam
|
|
Week 8
|
11/9
|
Mosaicing
|
|
Week 9
|
11/16
|
Lightness, Radiometry
|
|
Week 10
|
11/23
|
Photometric Stereo, Shape-from-Shading
|
|
Week 11
|
11/30
|
Stereo, Recognition
|
|
Week 12
|
12/7
|
Final Exam
|
Textbook
Robot Vision, by B.K.P. Horn, MIT Press, 1986. (ISBN:
0262081598)
Most of the lectures will follow this book. Although it is not required,
it is highly recommended. You can order this textbook from Drexel Bookstore.
Supplemental readings will be posted in Lectures.
The following is a list of general computer vision text books
recommended (but not required) for supplemental reading.
Computer Vision: A Modern Approach, by D.A. Forsyth
and J. Ponce, Prentice Hall, 2002. (ISBN: 0130851981)
A Guided Tour of Computer Vision, by V.S. Nalwa,
Addison-Wesley, 1993. (ISBN: 0201548534).
Computer Vision: Three-Dimensional Data from Images,
by R. Klette, K.Schluns, and A. Koschan, Springer Singapore, 1998.
(ISBN: 9813083719)
Assignments
Students will be assigned 3 multi-week individual projects. These
projects will bring all aspects of the learned material at each
stage. In each project, graduate students will be required to do
additional implementation. The first two of these projects will
also be competitions; students will vote for the top 3 artifacts in
class. Those who produced the top 3 artifacts will receive extra
credits according to their ranks. See Assignments for details.
You must be the sole original author of all assignments and
examination solutions in their entirety. As the university's
policy explains, penalties up to and including a failing grade for
the course with no opportunity to withdraw, will be given for
plagiarism, fabrication, or cheating*.
*The
standards for originality in a program are similar to those of
other written works. Programs by different authors show clear and
substantial differences as judged by most criteria, including but
not limited to: choice of variable and procedure names, line
spacing and indentation, choice of program structure, choice of
algorithms, ordering of modules, module design, and ordering and
choice of instructions. The original author of an assignment can
explain each detail and how they came to create it on their own.
Grading
Projects: 70% (23% x 3)
Exams: 30% (midterm and final ; closed book)
Assignments turned in up to one day late incur a 50% penalty;
assignments turned in more than one day late cannot be accepted and
receive a score of 0. Missed exams also receive a score of 0.
Make-up exams will only be allowed in extreme circumstances.
Communication
The instructor will disseminate important announcements by email
through the course mailing list, and also post these announcements
on the course web site. Also, the web site contains a time-line with
links to all information (lecture slides, assignments, etc.)
relevant to the course.
Policies
Attendance for lectures and exams is expected. In the case of
a school closing on an exam day, the exam will be given in the next
class period. The Drexel snow emergency information number is (215)
895-6358.
Academic honesty is essential. Cheating, academic
misconduct, plagiarism, and fabrication of any submitted material,
including both code and prose, are serious breaches of academic
integrity and will be dealt with accordingly. Violations will
result minimally in a grade of zero for the exam/assignment in
question, an additional reduction of one letter grade in the overall
course grade, and a report of the violation to the Drexel
administration; further penalties may apply to more serious and/or
repeat violations. Please refer to Drexel's official Academic
Honesty Policy for more information.
Lectures
Acknowledgments: A significant part of this course is
similar to the courses offered at Columbia (Shree Nayar), Univ. of
Washington (Steve Seitz) and CMU (Alyosha Efros). The instructor
thanks the instructors of these courses for the materials (slides,
content) used in this course. In addition, several photographs and
illustrations are borrowed from Internet sources. The instructor
thanks them all.
Permission to use/modify materials:
The instructor gladly gives permission to use and modify any of the
slides for academic and research purposes. Since a lot of the
material is borrowed from other sources, please acknowledge the
original sources too. Finally, since this is a continuously evolving
course, all suggestions and corrections (major, minor) are welcome!
Assignments
Project 1
Homography
Homography
In this project, you will implement a program to (1) rectify an
image (Image Rectification) (2) and superimpose one whole
image or a part of an image into another (Image Composite)
using manually selected correspondences on planar surfaces in the
images. Along the way, you will learn how to compute homographies
and how to use them to warp images.
Follow this link for
details. Make sure you read the page carefully. Those who need a
CS account and/or wish to use the lab camera, please contact the
instructor and TA by e-mail as soon as possible. Again, please go
to this page for
details.
Project 2
Mosaicing
In this project, you will implement a program to combine multiple
images into a panorama. The program has to automatically align the
input images by computing their relative motions and then blend
the resulting tile of images into a single seamless panorama.
Along the way, you will learn how to warp images into cylindrical
coordinates and compute translational motion between images using
a Gaussian pyramid.
Follow this link for
details. Make sure you read the page carefully. Please schedule
with the TA for the tripod, if necessary.
Project 3
3D Reconstruction
In this project, you will implement photometric stereo which is a
method to recover the 3D geometry of an object from multiple
images taken under varying illumination. The program will first
have to compute light directions and then, for each pixel,
estimate surface normals and recover depth information from the
surface normals.
Follow this
link for details. Make sure you read the page carefully.
Resources
|