17 December 1998 Fast edge map extraction from MPEG compressed video data for video parsing
Author Affiliations +
For the last few years, shot boundary detection has been recognized as an important research issue on video retrieval. Also, as a preliminary step for the task, it is essential to extract salient features from videos. Recently, it has become common to perform the two tasks in the compressed domain to alleviate their computational costs. In this paper, we propose a novel shot boundary detection technique, which uses two feature images, DC and edge image, which are extracted directly from MPEG compressed video. While a DC image can be easily obtained, edge image extraction usually requires a considerable computational burden. For fast edge image extraction, we suggest us of only a few AC coefficients of each DCT block, in motion compensated P-frames, B-frames, and I-frames. This drastically reduces the computational burden, compared to edge extraction in the spatial domain. In order to further reduce the computational burden, another edge image extraction technique is also suggested on the basis of AC prediction, using DC images. By using the edge energy diagram, obtained from edge images, and histograms from DC images, shot boundaries, such as abrupt transitions, fades, and dissolves are detected automatically. Simulation results show that the proposed techniques are fast and effective.
© (1998) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Byung Cheol Song and Jong Beom Ra "Fast edge map extraction from MPEG compressed video data for video parsing", Proc. SPIE 3656, Storage and Retrieval for Image and Video Databases VII, (17 December 1998); doi: 10.1117/12.333891; https://doi.org/10.1117/12.333891


Automatic generation of summaries for the Web
Proceedings of SPIE (December 17 2003)
Automatic home video abstraction using audio contents
Proceedings of SPIE (August 29 2002)
Detection of gradual scene changes for parsing of video data
Proceedings of SPIE (December 22 1997)
Online scene change detection of multicast (MBone) video
Proceedings of SPIE (October 04 1998)
Image feature meaning for automatic key-frame extraction
Proceedings of SPIE (December 17 2003)

Back to Top