A Novel Algorithm for Shot Boundary Detection and Segmentation of Broadcast News Videos

Abstract

News videos are an important means of conveying information. But the number of news broadcasted each day in a particular channel is very large and considering the number of news channels, it is a huge amount of information conveyed each day. Sorting of news videos plays a very important role in such cases to create a well-categorized database of news videos. Broadcast news video shot segmentation is one of the vital parts of sorting. This paper presents a novel algorithm that can detect the shot boundaries of news videos efficiently and perform shot segmentation accordingly. The algorithm adapts itself according to the input news video and hence is news channel independent.