Video Shot Detection using Texture Feature

Priyanka A R, Dr. Jharna Majumdar

Abstract: The multimedia data has the need for finding out some of the effective and robust methods mainly for storing the information in the database. Also, in its beginning step, accessing the content of the video necessitates the searching of video information. Hence, a huge amount of methods needs to be proposed in order to solve the main issues in the detection of shots. In this paper, first we demonstrate the Global and Local Pixel-wise Difference. Here, the main aim is to prove that the Local Pixel-wise difference method can accurately identify the exact position of change when a shot-transition occurs when it is compared to Global Pixel-wise difference. Secondly, we demonstrate the GLCM for Video-Shot-Detection in order to determine which feature gives maximum global difference and which feature gives maximum local difference when a shot transition occurs. Finally, we demonstrate the detection of the slow movement and jumping activity in a scene.

Keywords: Video Shot Detection, Global Pixel-Wise Difference, Local Pixel-Wise Difference, GLCM, Activity Tracking