J4 ›› 2012, Vol. 50 ›› Issue (06): 1185-1191.

• 计算机科学 • 上一篇    下一篇

一种基于镜头切换和文本检测的广告检测算法

孟丽1, 李月龙2, |陈刚1   

  1. 1. 军事交通学院 汽车指挥系, 天津 300161|2. 天津工业大学 计算机科学与软件学院, 天津 300387
  • 收稿日期:2012-05-21 出版日期:2012-11-26 发布日期:2012-11-26
  • 通讯作者: 孟丽 E-mail:mldream8@163.com

TV Advertisement Retrieval Based on Shot Change and Text Detection

MENG Li1, LI Yue long2, CHEN Gang1   

  1. 1. Department of Automobile Transport Command, Military Transportation University, Tianjin 300161, China;2. School of Computer Science and Software Engineering, Tianjin Polytechnic University, Tianjin 300387, China
  • Received:2012-05-21 Online:2012-11-26 Published:2012-11-26
  • Contact: MENG Li E-mail:mldream8@163.com

摘要:

针对广告的两个基本特征: 镜头切换较频繁和经常突显商标信息, 提出一种基于镜头切换和文本检测的广告检测算法. 该算法先对各相邻帧进行直方图差分, 再对剪切、 消融、 淡入淡出和划变4种常见的镜头切换方式进行分析和识别, 实现了镜头切换检测. 对于镜头切换频率较低的广告, 通过基于最大梯度差分的文本检测方法检测广告商标信息, 实现了广告自动检测系统. 实验结果表明, 对于约15 min的视频序列, 该检测算法可获得83.32%的正确率, 可初步满足广告检测的要求.

关键词: 广告检测; 镜头切换; 文本检测

Abstract:

Based on the  two basic features of advertisements: frequent shot changes and apparent brand information, the curve of histogram difference was obtained via histogram difference between two adjacent frames, and four common shot changes such as cut, dissolve, fade\|in/out and wipe were analyzed so as to  recognize shot change detection. As for those advertisements with low shot change frequencies, the text detection method based on maximum gradient difference was employed to extract the brand information. According to the above ideals and algorithms, a video advertisement retrieval system was designed and realized. A general precision rate of 8332% in about 15 min of video frames indicates the system could preliminarily satisfy the performance requirement according to video consecutive sequence.

Key words: advertisement retrieval, shot change, text detection

中图分类号: 

  • TP391