Video Summarization - An improved Framework using CNN

Authors

  • Suvarna Patil Department of Computer Engineering, SPPU, DYPCOE, Pune, Maharashtra, India
  • Dhanashree Phalke Department of Computer Engineering, SPPU, DYPCOE, Pune, Maharashtra, India

Keywords:

Video Summarization, Key Frames, Convolutional Neural Network, Static Summary, Dynamic Summary

Abstract

Summarization is an act of exhibiting the most significant thing which is used in every field like Video,
Audio and Text summarization. Video Summarization is a concise and meaningful representation of a video which is
mostly used in surveillance of data. Video summarization has been mainly preferred to enhance faster browsing of large
videos collection and more effective contents accessing and indexing. The proposed system which is summarizes the
video by using CNN technique. In this work, the input video is extracted number of frames, then applying Convolutional
Neural Network the extracted frames are experienced with the trained dataset for recognizing important frames. From
the sorted frames, the system generates the resultant summarized video. The proposed work generated more accurate
video summary. Various cricket videos are collected as a dataset.

Published

2019-06-25

How to Cite

Suvarna Patil, & Dhanashree Phalke. (2019). Video Summarization - An improved Framework using CNN. International Journal of Advance Engineering and Research Development (IJAERD), 6(6), 42–48. Retrieved from https://ijaerd.org/index.php/IJAERD/article/view/4275