摘要
研究了独立成分分析法(ICA)在视频序列处理中的应用,提出了一种新的基于ICA算法的视频字幕检测与定位方法,该算法采用了小波变换来提取独立分量.仿真实验结果表明,与传统方法相比,该方法能够有效消除背景信息,提高字幕定位的精度,具有更强的鲁棒性.
Independent Component Analysis(ICA)is applied to video frame sequence processing.A novel scheme for the detection and location of the text in video based on ICA algorithm is proposed.The new algorithm uses wavelet transformation to extract independent components.The simulation results prove that compared with the conditional algorithm,the algorithm can eliminate the background information and improve the precision of the caption location.Thus the algorithm is stronger robustness.
出处
《西安石油大学学报(自然科学版)》
CAS
北大核心
2011年第3期100-103,12,共4页
Journal of Xi’an Shiyou University(Natural Science Edition)
基金
陕西省电子信息系统综合集成重点实验室基金资助(编号:201107Y16)
关键词
视频字幕
字幕检测
字幕定位
独立成分分析
text in video
caption detection
caption location
independent component analysis