Automated Scene Matching in Movies [chapter]

F. Schaffalitzky, A. Zisserman
2002 Lecture Notes in Computer Science  
We describe progress in matching shots which are images of the same 3D scene in a film. The problem is hard because the camera viewpoint may change substantially between shots, with consequent changes in the imaged appearance of the scene due to foreshortening, scale changes and partial occlusion. We demonstrate that wide baseline matching techniques can be successfully employed for this task by matching key frames between shots. The wide baseline method represents each frame by a set of
more » ... nt invariant local feature vectors. The local spatial support of the features means that segmentation of the frame (e.g. into foreground/background) is not required, and partial occlusion is tolerated. Results of matching shots for a number of different scene types are illustrated on a commercial film. Invariant descriptors for multiview matching In this section we describe the invariant descriptors which facilitate multiple view matches, i.e. point correspondences over multiple images.
doi:10.1007/3-540-45479-9_20 fatcat:dnyrqfo2jjektoho6oi5ex2hfu