The goal of this laboratory class is to gain experience in implementing robotic visual perception algorithms on modern computer hardware. Students will be able to choose a specific project which they will implement in C/C++ during the class, with results presented at the end of the class.After this class, students will have gained substantial insights on how to implement image processing and visual perception algorithms using one of the most common programming languages used on embedded robotics compute hardware.

This practical class builds on the materials taught in the “Topics from Robotics: Visual Perception” class. Additional material will be presented interactively during the class according to the selected project. Students will form small groups to implement their project. They will be provided with a programming environment (Laboratory computer) and example data. Depending on the number of students subscribed to the class and the number of available computers, students might be required to share a computer.

Implementation of a selected visual perception project in C/C++Project examples are:Camera modellingPanoramas/Image registrationFeature tracking and motion estimation3D perceptionVisual odometryCamera simulation/ renderingEtc.

Good programming skills in C/C++Good knowledge of Mathematics (linear algebra)Intermediate knowledge in Computer Vision Basic knowledge in RoboticsWhile not required, it is strongly suggested to also subscribe to the “Topics in Robotics: Visual Perception” class, that provides an understanding of the fundamentals required for the implementation of projects in this class.

Books:Richard Szeliski: Computer Vision: Algorithms and Applications, 2nd ed. Paper or online.David A. Forsyth: Computer vision: A modern approach. Paper or online.Link auf weitere Informationenhttps://campus.aau.at/studium/course/112865