Abstract
Due to the advancement in multimedia technologies and wide spread usage of internet facilities; there is rapid increase in availability of video data. More specifically, enormous collections of Medical videos are available which has its applications in various aspects like medical imaging, medical diagnostics, training the medical professionals, medical research and education. Due to abundant availability of information in the form of videos, it needs an efficient and automatic technique to manage, analyse, index, access and retrieve the information from the repository. The aim of this paper is to extract good visual content representatives – Summary of keyframes. In order to achieve this, the authors propose a new method for video shot segmentation which in turn leads to extraction of better keyframes as representative for summary. The proposed method is experimented and evaluated using publically available medical videos. As a result, better precision and recall is obtained for shot detection when compared to that of the recent related methods. Evaluation of video summary is done using fidelity measure and compression ratio.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call