TV Commercial Detection in News Videos
Date Issued
2004
Date
2004
Author(s)
Yeh, Jen-Hao
DOI
en-US
Abstract
Detection of the commercials in TV program videos is hard because the diversity of them puts up a rather high barrier to find an appropriate model. After some studies of existing commercial detection works, the production rule of commercial and some observations, we try to deal with this problem through a top-down approach. We select some commercial features by conjecturing the thoughts of both commercial producers and the audiences. By combining the commercial features with a video scene detection mechanism, a two-level commercial detection scheme is proposed. Human intervention is also included if better results are required. Human intervention also let users cutting commercial more effectively. The experiments show good results of proposed scheme on detecting commercials in news programs videos
Subjects
電視廣告偵測
新聞影片分析
commercial detection
news video analysis
Type
thesis
