While computers outperform humans in many areas, e.g., doing arithmetic, searching databases, and even playing chess, other problems remain notoriously difficult. For example, a two-year-old easily beats a computer at recognizing people and identifying objects in images. In CS 453 we study the problem of extracting information from digital images and movies, covering topics such as edge and motion detection, stereo vision, and object recognition. Along the way we will discover why such tasks are hard, and gain experience with analyzing real, noise-contaminated data.


Prerequisites: CS 202, MATH 200

Computer vision is my main area of research. I find it fascinating because image interpretation seems effortless for humans, yet solving these problems by computer is extremely hard. Interestingly, some of the greatest progress has been made by applying mathematical techniques, which can be an eye-opening experience for students.
-- Prof. Scharstein