Loading...
Please wait, while we are loading the content...
Similar Documents
Automatic Temporal Segmentation for Content-Based Video Coding Technical Report
| Content Provider | Semantic Scholar |
|---|---|
| Author | Habili, Nariman Moini, Alireza |
| Abstract | As video coding algorithms and video coding standards become more reliant on content-based manipulation, there is a greater need for robust object segmentation techniques. In this paper we present a temporal segmentation technique based on significance testing, connected components analysis and mathematical morphology. We first derive the difference image between two consecutive images of a video sequence and form a histogram. The image noise, which is assumed to follow a zero-mean Gaussian distribution, is therefore reflected in the histogram and we use a fitting technique to find the variance of this noise. We obtain change detection thresholds based on suitable false alarm probabilities. Connected components analysis is then utilized to remove any residual noise, and the morphological close operator is applied to obtain the segmentation mask. Results indicate that our technique is robust and also immune to noise. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.eleceng.adelaide.edu.au/Personal/nhabili/techrep.pdf |
| Alternate Webpage(s) | https://www.researchgate.net/profile/Alireza_Moini/publication/228790906_Automatic_temporal_segmentation_for_content-based_video_coding/links/09e415148d1464f579000000.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
