公开(公告)号：EP2173090B1

公开(公告)日：2014-11-05

申请号：EP09171884.1

申请日：2009-09-30

申请人： Thomson Licensing

发明人： le Meur, Olivier ,  Chamaret, Christel ,  Chevet, Jean-Claude

IPC分类号： G06T1/00

CPC分类号： H04N1/32203 ,  G06T11/60 ,  H04N1/32229 ,  H04N1/32325

公开(公告)号：EP2175654A1

公开(公告)日：2010-04-14

申请号：EP09171920.3

申请日：2009-10-01

申请人： Thomson Licensing

发明人： le Meur, Olivier ,  Ninassi, Alexandre ,  Urban, Fabrice

IPC分类号： H04N7/24

CPC分类号： G11B27/036 ,  H04N21/23418 ,  H04N21/23424 ,  H04N21/44016 ,  H04N21/812 ,  H04N21/8455

摘要： The present invention relates to a method for processing pictures intended to insert an advertising clip at a point, called insertion, between two pictures of a sequence of video pictures, called video sequence, comprising the following steps:
- generating (E1) a salience map representing the salience of the video sequence preceding the insertion point,
- generating (E2), for each advertising clip of a set of advertising clips, a salience map,
- determining (E3), for each advertising clip of said set of advertising clips, a degree of similarity between the salience map of the video sequence and the salience map of said advertising clip, said degree of similarity being representative of the comparison between the location of the salience zones on both said maps,
- selecting (E4), among said set of advertising clips, the advertising clip having the highest degree of similarity, and
- inserting (E5) the advertising clip selected into the video sequence at the insertion point.

摘要翻译： 本发明涉及一种用于处理图像的方法，所述图像旨在在称为视频序列的视频图像序列的两个图像之间的称为插入的点处插入广告剪辑，包括以下步骤： - 生成（E1）显着图 代表插入点之前的视频序列的显着性， - 针对一组广告片段的每个广告片段生成（E2），对于所述一组广告片段的每个广告片段的显着性映射，确定（E3） 所述视频序列的显着性图和所述广告片的显着性图之间的相似程度，所述相似度表示所述两个所述地图上所述显着区域的位置之间的比较， - 选择（E4） 广告片段，具有最高相似度的广告片段;以及 - 插入（E5）所选择的广告片段到插入点处的视频序列。

公开(公告)号：EP2173090A1

公开(公告)日：2010-04-07

申请号：EP09171884.1

申请日：2009-09-30

申请人： Thomson Licensing

发明人： le Meur, Olivier ,  Chamaret, Christel ,  Chevet, Jean-Claude

IPC分类号： H04N5/272 ,  G06T7/00

CPC分类号： H04N1/32203 ,  G06T11/60 ,  H04N1/32229 ,  H04N1/32325

摘要： The present invention relates to a method for insertion of layout, such as a logo or a advertising message, in at least on image of a sequence of source images. It comprises the following steps:
a) analysis (100, 110, 120; 200, 210) of the video content of potential insertion areas of the source image and the layout or video content of potential insertion areas of the source image after preliminary insertion of said layout in said potential insertion areas in such a way to generate video content information,
b) selection (130; 220, 230), among said potential insertion areas, of an insertion area according to said video content information, and
c) insertion (140; 240) of said layout in the selected insertion area.

摘要翻译： 本发明涉及一种用于在源图像序列的至少图像上插入诸如徽标或广告消息的布局的方法。 它包括以下步骤：a）源图像的潜在插入区域的视频内容的分析（100,110,120,2200,210）以及在初步插入之后源图像的潜在插入区域的布局或视频内容 在所述潜在插入区域中以这样一种方式生成视频内容信息的方式进行布局; b）根据所述视频内容信息在插入区域的所述潜在插入区域中选择（130; 220,230），以及c）插入（ 140; 240）所选择的插入区域中的布局。