Back close

An evolutionary computing approach for solving key frame extraction problem in video analytics

Publication Type : Conference Paper

Publisher : Communication and Signal Processing (ICCSP), 2017 International Conference on

Source : Communication and Signal Processing (ICCSP), 2017 International Conference on, IEEE, Chennai, India (2017)

Url : https://ieeexplore.ieee.org/abstract/document/8286663/

Campus : Coimbatore

School : School of Engineering

Department : Computer Science

Year : 2017

Abstract : Key frame extraction is a very important task in video analytics. The extracted key frames are used for video summarization and event detection for the given videos. This paper proposes a new algorithmic framework which combines the conventional SSIM (Structural Similarity Index) approach with the state-of-the-art Differential Evolution (DE) algorithm for key frame extraction from video data. The proposed algorithm is implemented to extract key frames from three different videos and its performance is compared with the conventional approach using the average SSIM values. The results show that DE based approach gives comparatively better results when compared to the naive SSIM approach.

Cite this Research Publication : K. Thomas Abraham, Ashwin, M., Sundar, D., Ashoor, T., and Dr. Jeyakumar G., “An evolutionary computing approach for solving key frame extraction problem in video analytics”, in Communication and Signal Processing (ICCSP), 2017 International Conference on, Chennai, India, 2017.

Admissions Apply Now