This course will present standard theories and techniques for the processing of still images and videos. From a definition of the problem of visual perception and some elements of answer on how it is solved by our brain, we will move on to explain theories for the extraction of significant structures in images. The tasks of object and pattern recognition, and tracking will be discussed.
The course will be concluded by a group project. The project is intended to make the students apply the theory learned during the course on practical problem.
The general task is to create a marker and to locate this marker in an image sequence performing rather smooth translations as well as rotations as precisely and reliably as possible. For that you should make use of different visual modalities (colour and shape information) as well as a temporal regularization (for example in terms of a Kalman filter).
You can download example sequences