This course focuses on descriptions of objects at two main levels of abstraction: Segmented images-images organized into subimages that are likely to correspond to interesting objects, and Geometric structures-quantitative models of image and world structures. Basic techniques for image processing and feature extraction are covered in lectures; topics include: image formation, edge detection, region growing, and shape description. The higher-level more experimental image analysis techniques, such as video image sequence analysis, are covered by selected presentations and projects.

During the first part of the semester a sequence of computer labs will provide experience in the software tools that are important for computer vision applications. In the last part of the semester, students will do an in-depth project, which builds on their lab experience, on a topic that they select. Typically, a project will involve exploring a state-of-the-art technique that will include research literature review.

ECE5470 is a first course on learning about how computers can see; that is, interpret the multidimensional signals provided by imaging sensors. Computer vision has a very wide range of applications from medical diagnosis to seeing robots, from particle physics to geological surveying. Wherever images play an important role in understanding a problem, there is a potential application for computer vision. The objective of this course is to provide students with an understanding of the fundamental methods and an appreciation for the state of the art and the potential of computer vision.