B. Gunsel and A. M. Tekalp. Content-based video abstraction. In IEEE International Conference on Image Processing, volume 3, pages 128--132, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....to summarize movies semantically and apply similar techniques. In the attempts to structure films into logical components, only information about the contents of the film can help, or, as in [9] by the analysis of frame information. Further applications of video summarizations can be found in [6] and [15] A number of computer vision techniques and hiddenmarkov models can be applied to 2D images to describe a sequence[2] However, good results are only achievable under a controlled lab environment. Eickeler et al. 4] try a similar approach, but limit it to recognizing human gestures. ....

Bilge Gunsel and Murat Tekalp. Content-based video abstraction. In IEEE Int'l Conference on Image Processing - ICIP98, Chicago, Illinois, volume 3, pages 128-132, 1998.

....so as to facilitate further processing such as indexing, key frame extraction and shot content analysis. Techniques including colour histogram, motion and frequency analysis are used to determine when shot changes have occurred, representative key frames are then extracted from individual shots [6, 7, 5]. Video skimming introduces the use of audio and language analysis to generate condensed representations of the original material [8] In our approach we are not concerned with shot segmentation, just that the visual abstraction generates a small number of informative key frames. This then ....

B. Gunsel and A. M. Tekalp, Content-Based Video Abstraction, IEEE International Conference on Image Processing, Chicago, Il, Oct., 1998.

.... been used for shot boundary detection are pair wise frame differences [1] statistical differences [2] histogram comparisons [2] 3] edge differences [2] and different source integration techniques [4] 6] most of which involve the a priori or adaptive determination of an appropriate threshold [7], 3] This work has been supported by the European Project RTN1 199900177 MOUMIR Instead of employing thresholding methods in order to identify shot cuts, the problem may be reformulated as a pure classification problem concerning two main clusters: one cluster representing shot boundaries and ....

Bilge Gunsel and A. Murat Tekalp, "Content-based video abstraction, " in Proc. of 1998 IEEE International Conference on Image Processing, Chicago, Illinois, USA, 1998.

....keyframes. An easy and straightforward approach is to select one keyframe per shot, usually the n th frame for a fixed number n. To capture the concept of a shot with a great deal of motion, methods for sequentially selecting multiple keyframes have also been proposed based on frame di#erencing [21, 34]. Other methods for selecting multiple keyframes include the use of motion information [4] and clustering approaches [5, 22] 3.1 Shot Representation Using the Shot Tree This section describes a novel way to represent a shot based on a tree structure, known as a shot tree, formed by clustering ....

Bilge Gunsel and A. Murat Tekalp, "Content-based video abstraction," in Proceedings of IEEE Int'l Conference on Image Processing, Chicago, IL, October 4-7 1998.

No context found.

B. Gunsel and A. M. Tekalp. Content-based video abstraction. In IEEE International Conference on Image Processing, volume 3, pages 128--132, 1998.

No context found.

Bilge Gunsel and A. Murat Tekalp, "Content-based video abstraction," in Proceedings of IEEE Int'l Conference on Image Processing, Chicago, IL, October 4-7 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC