3D Manipulation of 2D Images
Dr. Peter L. Cho, Alexandru N. Vasile
Tue Jan 10, Fri Jan 13, Tue Jan 17, Fri Jan 20, 12-01:30pm, 56-154
Enrollment limited: advance sign up required (see contact below)
Signup by: 06-Jan-2012
Limited to 20 participants.
Participants requested to attend all sessions (non-series)
Prereq: MATLAB and/or linux experience helpful. See description.
Billions of digital images are being collected each year. But current hardware abilities for gathering electronic pictures far exceed conventional software capacities for organizing and searching these data. In this course, we survey recent advances in computer vision which utilize 3D geometry to manipulate 2D imagery. As we'll see, geometry-based approaches to image processing coupled with internet-scale computing imply many neat, new applications.
Each class will begin with a theoretical overview and end with a computer lab. The primary topics for the 4 sessions are multi-view geometry, automatic feature matching, panorama formation and 3D reconstruction. During the computer labs, students will work with open-source and commercial software in order to calibrate cameras, extract SIFT features, form mosaics and generate 3D point clouds from multiple photos.
This work is sponsored by the Department of the Air Force under Air Force Contract #FA8721-05-C-0002. Opinions, interpretations, conclusions and recommendations are those of the authors and are not necessarily endorsed by the United States Government.
Contact: Dr. Peter L. Cho, LIN-S3-300, (781) 981-2802, cho@ll.mit.edu
Sponsor: Lincoln Laboratory
Cosponsor: Electrical Engineering and Computer Science
Latest update: 15-Dec-2011
|
|