Automatic partitioning of full-motion video
Multimedia Systems
The nature of statistical learning theory
The nature of statistical learning theory
A unified shot boundary detection framework based on graph partition model
Proceedings of the 13th annual ACM international conference on Multimedia
Evaluation campaigns and TRECVid
MIR '06 Proceedings of the 8th ACM international workshop on Multimedia information retrieval
Analysis of Historical Artistic Documentaries
WIAMIS '08 Proceedings of the 2008 Ninth International Workshop on Image Analysis for Multimedia Interactive Services
Video shot boundary detection: Seven years of TRECVid activity
Computer Vision and Image Understanding
A unified model for techniques on video-shot transition detection
IEEE Transactions on Multimedia
Video Segmentation via Temporal Pattern Classification
IEEE Transactions on Multimedia
Shot-boundary detection: unraveled and resolved?
IEEE Transactions on Circuits and Systems for Video Technology
A Formal Study of Shot Boundary Detection
IEEE Transactions on Circuits and Systems for Video Technology
Hi-index | 0.00 |
The segmentation of films and videos into shots requires the detection of gradual transitions such as dissolves and fades. There are two types of approaches: unified approaches, that is, one detector for all gradual transition types, and approaches that use specialized detectors for each gradual transition type. We present an overview on existing methods and extend an existing unified approach for the detection of gradual transitions in historic material. In an experimental study, we evaluate the proposed approach on complex and low-quality historic material as well as on contemporary material from the TRECVid evaluation. Additionally, we investigate different features, feature combinations, and fusion strategies. We observe that the historic material requires the use of texture features, in contrast to the contemporary material that, in most of the cases, requires the use of color and luminance features.