An adaptive novel feature based approach for automatic video shot boundary detection

摘要

The video shot boundary detection is an area, where many techniques and methods have been suggested and have claimed to perform consistently on the videos like television broadcasting of different sports events, talk shows & news telecast. The movies contain a large amount of visual information varying in arbitrarymanner; it requires a highly vigorousprocedure for automatic shot boundary detection. The main goal of this research is to develop an adaptive & robust algorithm for segmentation & classification of movie videos. In this paper, we described a novel approach for shot boundary detection which is totally adaptive & changes the threshold value as per the information change. We have defined the novel feature which can uniquely define color as well as edge & frequency details of a frame. It is able to detect not only abrupt transitions like hard cuts but also special effects like wipes, fades, and dissolves as well in different movies.