A. Juozapavicius "Computer Vision"

Catalog description:

This course (project) has a goal to give an experience for students to design and implement a computer vision system. It will cover fundamental algorithms of computer vision which underlie many applications, like robotics, remote sensing, etc. Students may choose an application according to their interests.

Current texts:

• Mikhail J. Atallah (editor). Algorithms and Theory of Computation Handbook, CRC Press LLC, London, New York, 1999
• Bernd Jahne, Horst Haussecker. Computer Vision and Applications, Academic Press, 2000

Goals:

The course (project) is intended as an introduction and survey of the field of Computer Vision. The emphasis will be on fundamental techniques which underlie many computer vision research and applications (including robotics, molecular biology, satellite images, etc.). The experience in design and implementation of computer vision systems will be gained.

Content:

• Architecture and components of computer vision (image processing for black-white, colour, range and thermal images, extraction of symbolic and structural features, identification and location of objects);
• image connectivity analysis (algorithms of connected components, including identification of connected components in binary images, as well as in greyscale or colour images);
• the Hough transform (algorithms of line detection, as well as detection of other parametric curves);
• model-based object recognition (matching sets of feature points, matching contours of planar shapes, matching relational description of shapes);
• analyis of the design and implementation issues.

Typical requirements:

Courses in image processing, logic and programming, discrete mathematics are necessary

Helpful background:

Programming in C++ or graphics programming tools are desirable