<<
| Signature | DA 867 |
| Author | Thilo Borgmann |
| Title | Video Analysis for Camera Motion Classification |
| Tutor | Dipl.-Ing. Andreas Krutz |
| Professor | Dr.-Ing. Thomas Sikora |
| Date | 30.01.2009 |
| Abstract | To know the type of camera motion related to a video sequence is an important ascpect in the fields of computer vision and video processing. Many algorithms rely on assumptions about this camera motion type. Therefore, to classify this motion is an important task to ensure the applicableness of certain algorithms. In this thesis, a complete algorithm is developed which is capable of classifying the camera motion type of a given sequence. This algotithm uses uncalibrated camera models and unsupervised classification to achieve the this classification. |
| Key words | video segmentation, camera motion, classification |
BibTeX