Static Video Summarization Using Artificial Bee Colony optimization
This paper proposes an Artificial Bee Colony optimization based static video summarization technique that represents the synopsis of a video by the most vital frames present in it. First the pixel clusters or Regions of Interest of the video frames which capture the most significant content variatio...
Saved in:
Published in | 2018 IEEE Symposium Series on Computational Intelligence (SSCI) pp. 777 - 784 |
---|---|
Main Authors | , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.11.2018
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | This paper proposes an Artificial Bee Colony optimization based static video summarization technique that represents the synopsis of a video by the most vital frames present in it. First the pixel clusters or Regions of Interest of the video frames which capture the most significant content variations are identified. A novel set of features measured in terms of the average hue values of each of these regions is then used to characterize the frames. Based on these features, the sequence of frames of the video is divided into segments by the Artificial Bee Colony optimization algorithm. The lengths of the segments are so optimized that all the frames of a particular segment have similar features, whereas, the middle frames of different segments are significantly different from each other. These middle frames are treated as the key-frames of the video. Any redundancy present in the selected key-frames is eliminated by comparing their hue histograms. The proposed work is validated on the publicly availableSumMe dataset and on a set of randomly selected web videos. The obtained results and the comparisons with the state-of-the-art methods conclude that the proposed work is well applicable for summarizing videos of different genres. |
---|---|
DOI: | 10.1109/SSCI.2018.8628784 |