Video summarization provides condensed and succinct representations of the content of a video stream. A static storyboard summarization approach based on robust low-rank subspace segmentation is proposed in this paper. Firstly, video frames are represented as multi-dimensional vectors, and then embedded into a group of affine subspaces using low-rank representation according to the content similarity of the frames in the same subspace. Secondly, a series of subspaces are segmented based on the Normalized Cuts algorithm. The video summary is finally generated by choosing key frames from the significant subspaces and ranking these key frames in temporal order. The experimental results demonstrate that the proposed summarization algorithm can produce crucial key frames and effectively reduce the visual content redundancy in summary comparing with the conventional approaches. © Springer-Verlag Berlin Heidelberg 2013.
CITATION STYLE
Tu, Z., Sun, D., & Luo, B. (2013). Video Summarization by Robust Low-Rank Subspace Segmentation. Advances in Intelligent Systems and Computing, 212, 929–937. https://doi.org/10.1007/978-3-642-37502-6_109
Mendeley helps you to discover research relevant for your work.