Comparative Performance of Shot Change Detection Techniques in MPEG Stream | ||
| Journal of University of Babylon | ||
| Article 1, Volume 24, Issue 3, April 2016, Pages 609-621 | ||
| Author | ||
| Ruaa Hussam Ali Al-Mallah | ||
| Abstract | ||
| This paper deals with the detection of shot changes in MPEG video stream. There are two main types of shot changes in a video. The first occurred abruptly, while the second type occurred gradually. This research is focused on the abrupt shot change detection. There are many techniques to detect a sudden shot change in video. In this research, the famous three techniques are used which are pixel difference, XOR operator, and Edge Change Ratio (ECR). Two new algorithms proposed with these techniques to improve the behaviors and to find the best results of these techniques; the first algorithm depends on grayscale levels and the second algorithm depends on edge detection operation by Soble operator. Later on, these two algorithms are applied with the three techniques in two videos, which have abrupt shot change to measure maximum difference values between each two frames. The first algorithm with pixel difference technique gives an accurate result 100%, while the perfect results are given with second algorithm applied with ECR and XOR techniques | ||
| Keywords | ||
| Abrupt shot detection; Edge Detection; Cut; Shot Boundary Detection; Edge Change ratio | ||
|
Statistics Article View: 125 PDF Download: 96 |
||