engleski

Analysis of hollywood’s film production using background subtraction visual activity index based on computer vision algorithm

The amount of available video content is increasing every day. Apart from the classic channels of video content distribution, such as television and satellite broadcasting, more and more video content is distributed over the Internet. For some decades, some authors have analysed the properties of the available video content and noted certain trends, and above all, a growing visual activity. This paper describes two measures of visual activity (Visual Activity Index and Average Shot Length), as well as papers that analyse their influence. Also, a new measure based on computer vision algorithms is proposed, and this measure is compared with Visual Activity Index. This proposed measure is more resistant to noise than Visual Activity Index, and this insensitivity is based on selected background subtraction algorithm that has this feature in its base for the proposed measure. The proposed measure called the Background Subtraction Visual Activity Index (BGSVAI) is used at the end of the paper to measure the visual activity of 50 films, which received the Award for Best Picture from the Academy of Motion Picture Arts and Sciences from 1965 to 2014. Also, the visual activity of these films was analysed in the mentioned period.

Visual activity, Background subtraction Algorithms, VAI, BGSVAI

nije evidentirano