Adaptive Methods for Motion Characterization and Segmentation of MPEG Compressed Frame Sequences
A fast and accurate method for scene change detection and classification of camera motion effects in MPEG compressed video is proposed. The method relies on adaptive threshold estimation and on the analysis and combination of various types of video features derived from motion and intensity information. This analysis is also applied for cleaning-up false shot boundaries due to camera motion effects. Two techniques for adaptive threshold estimation are also proposed and evaluated.
Unable to display preview. Download preview PDF.
- 2.Lienhart, R.: Comparison of Automatic Shot Boundary Detection Algorithms. Image and Video Processing (1999)Google Scholar
- 3.Truong, B.T., Dorai, C., Venkatesh, S.: New Enhancements to Cut, Fade, and Dissolve Detection Processes in Video Segmentation. In: ACM Multimedia Conference, California, USA (2000)Google Scholar
- 4.Kobla, V., et al.: Compressed Domain Video Indexing Techniques Using DCT amd Motion Vector Information in MPEG Video. Storage and Retrieval for Image and Video Databases (1997) Google Scholar
- 5.Yusoff, Y., Christmas, W., Kittler, J.: Video Shot Cut Detection Using Adaptive Thresholding. In: British Machine Vision Conference, Bristol, U.K. (2000)Google Scholar
- 6.Yeo, B.L., Liu, B.: Rapid Scene Analysis on Compressed Video. IEEE Trans. on Circuits and Systems for Video Technology 6, 533–544 (1995)Google Scholar