INSTRUCTOR
Prof. Gabriel Taubinemail: taubin@brown.edu
TEACHING ASSISTANT
Daniel Morenoemail: Daniel_Moreno@brown.edu
COURSE DESCRIPTION
The course provides an introduction to modern Image Processing algorithms, including Interactive Image Editing algorithms, Feature extraction algorithms, Feature-based image registration, and stitching algorithms, Featured-based recognition systems, Stereo and Multi-View Stereo algorithms, and various applications.
WHO SHOULD TAKE THE COURSE / PREREQUISITES
This course is aimed at senior/junior undergraduate students and graduate students from Engineering, Computer Science, Applied Mathematics, Physics, Cognitive Science and Neuroscience, with some knowledge of Vector Calculus, Linear Algebra,and Data structures, and exposure of computer programming. No prior knowledge of Image Processing or Computer Vision algorithms is assumed.
OFFICE HOURS
By appointment.
TEXTBOOK
There is no required book, although Richard Szeliski's "Computer Vision, Algorithms and Applications" is a good reference. We will download and read selected papers or chapters of ebooks available through the library. Additional materials will be handed out and will be available for download from this web site. All lecture slides will be available for download from this web site as well, after class.
GRADING
The course evaluation will be is based on biweekly programming assignments, and a final project. There will not be homework assignments other than the programming assignments and the final roject, and there will be neither midterms nor final exam. Programming will in Matlab, OpenCV and Java. Final projects implemented in mobile platforms will be encouraged. As new concepts and techniques are introduced in the lectures, the programming assignments will expose the students to image processing and analysis algorithms of increasing complexity. In the final project the students will either reproduce a state-of-the-art algorithm selected from the recent literature, or work on a novel problem. The final project requires a working implementation, a publication-quality report, and a class presentation in a conference setting. The goal for a novel final project is to actually submit the resulting work for publication.
Class attendance and participation in the class discussions is mandatory, and will contribute to the final grade. The final project will contribute 30% to the final grade, the programming assignments will contribute 60% to the final grade. Class participation will contribute the remaining 10%.
TENTATIVE TOPICS