Structure and event mining in sports video with efficient mosaic
Multimedia Tools and Applications
Single view geometry and active camera networks made easy
MiFor '09 Proceedings of the First ACM workshop on Multimedia in forensics
Towards obtaining an ideal real time panoramic video
ISVC'06 Proceedings of the Second international conference on Advances in Visual Computing - Volume Part II
Intelligent multi-camera video surveillance: A review
Pattern Recognition Letters
| Hi-index | 0.00 |
We address the problem of constructing mosaics fromvideo sequences taken by rotating cameras. In particular,we investigate the widespread case where the scene is notonly static but may also contain large dynamic areas, inducedby moving or deforming objects. Most of the existingtechniques fail to produce reliable results on such video sequences.For such alignment purposes, two classes of techniquesmay be used: feature-based and direct methods. We deriveboth of them in a unifiedstatistical manner and proposean integrated framework to construct what we call motionpanoramas, based on a mixed feature-based and direct approach.Experimental results are provided on large image sequences.In particular, we consider sport videos where themoving and deforming athlet is visible in every frame of thesequence, thereby making tricky the alignment task.