CS6850 Image Understanding
Homework assignment notes/hints page
A Style Guide is available
on-line to assist you in determining the correct style for your
programs. You are required to follow the guidelines in all programs
you turn in for the course. Failure to follow the guidelines may
result in a significantly lower grade on an assignment.
- Instructor: David M. Chelberg (Press here to email)
- Class Number 13497
- Office: Stocker 322B
- Office Hours: Mon, Wed, Fri 3pm-4pm and by email
appointment. Feel free to just stop by as well.
- Lecture Notes
- will be available from prime in the directory:
~cs685/lectures. They are stored as compressed postscript
files. To print, you must first uncompress using the command
gunzip, then send to a postscript printer (using lp). Detailed instructions for those new to
- To provide students with a comprehensive overview of the
principles of image understanding. To become acquainted with many
current techniques in image understanding.
- Proficiency in C++, data structures and algorithms, and
- Required Texts:
- Computer Vision, by Linda G. Shapiro, and George C. Stockman,
Prentice-Hall, Upper Saddle River, New Jersey, 2001. ISBN
- Course Outline:
- Topics include:
- Introduction to basic concepts in image understanding.
- Mathematical foundations of geometric transformations
- Low-level image analysis methods, including:
- image formation
- camera calibration
- edge detection
- feature detection
- Color image segmentation, and processing.
- Methods for inferring three-dimensional information from
2D images including shape from shading and depth from stereo.
- Three-dimensional object modeling and recognition.
- Students are expected to spend a substantial amount of time
outside of class working exercises in the book, and programming
- Examination schedule:
- There will be two midterm exams
approximately weeks 5 and 10. The final exam is replaced by a final
project. Final project presentations will take place at the
time/date of the final.
- Attendance Policy:
- Students are strongly encouraged to attend all classes, but
attendance is not required. Class attendance will not be used
in the final determination of grades. Students miss classes at
their own risk. Students are required to attend class during
the midterm and final exam unless prior arrangements have been
- Academic dishonesty:
- Students are expected to turn in only their own work with proper
documentation. Anything else will result in an F for the exam,
project or program, and possibly an F for the course, or even
dismissal from the University. This means NO WORKING IN GROUPS
(unless expressly allowed), and NO SHARING CODE.
- Sensor Animation
CCD vs CMOS explained with animations
- Medial Axis Animation
Dinosaurs, and medial axis computation
- Code Resources
Library Sourceforge Project
- Morphology Links
- Grayscale Morphology with Demo
- Hypermedia Image Processing Java Code demos
- CVonline: Morphological Transformations
- Robotics Links
Flex picker robots -- very fast
- Misc Resources
method for histogram based thresholding.
Lecture notes on histograms from Prof Emmanuel Agu, Worcester
moments, including how to calculate the orientation of a region
from the moments.
Eigen values and vectors of a 2x2 matrix.
Facial Recognition Tested
The computer vision homepage (lots of good links)
- USC Bibliography of Computer Vision
- A good free on-line text book of computer vision.
- CV Online General
Overview of Computer Vision with lots of resources.
- Ballard and Brown Computer Vision Text on-line (classic computer vision book)
- Intelligence: The Eye, the Brain and the Computer, M. A. Fischler, O. Firschein, Addison-Wesley, 1987, ISBN: 0201120011
Model Recognition From Stereoscopic Cues, J.E. W. Mayhew,
J.P. Frisby; MIT Press, 1991, ISBN 0-262-13243-5
A good review of probability for Image Understanding/Computer
- A good site
with an overview of the Canny edge detector, and an interactive
University Class Note site (good source for camera calibration
Canny's Edge Detection Paper 1986 (from TPAMI).
Alternate site for Canny's Edge Detection Paper 1986 (from
Good video about tristimulus theory of color, but funny, and other
random stuff as well.
David M. Chelberg <email@example.com>
last-modified: Thu Oct 23 11:33:49 2014