CS 294-6: Computational Imaging


This is a graduate advanced topics class offered in the Spring 2002 semester. We will study various methods for turning computers into image forming devices.

In many fields, compute-intensive reconstruction methods produce images from raw data that may be considered to be in a coded form. Often the image construction involves inversion of an operation that is simple in the forward direction. Since computing is becoming cheaper and faster, replacing optics, or electromagnetic apparatus, or electronics with computing can be a driving force in some such image forming applications.

A basic understanding of image processing and/or machine vision will be assumed. Familiarity with concepts of linear systems, convolution, and Fourier transforms will be handy.

There will be two lectures a week, and I hope to encourage discussion and student participation. There will be some papers to read, although in some of the areas to be covered there isn't that much written yet. There may be homework problems from time to time.

Course description

CS 294-6: Computational Imaging.
Advanced topics class on image formation by computational means.
Examples to be covered include:
Coded Aperture Imaging
Synthetic Aperture Microscopy
Diaphanography or "Optical Tomography"
Exact Cone Beam Reconstruction
May cover other topics according to student interest.
For additional details click here.
(2 units)

Instructor: Berthold K.P. Horn
Time: MW 10:30AM-12:00PM
Location: 310 Soda
Web page: http://www.ai.mit.edu/people/bkph/courses/cs294-6/compimag.html

Berthold K.P. Horn bkph@ai.mit.edu, http://www.ai.mit.edu/people/bkph/.