【摘要】 该文详细的介绍了独立元分析方法 (IndependentComponentAnalysis ,ICA)的基本背景、原理 ,以及各种独立元分析算法及其特点。同时考虑到在视频分割中 ,视频图像的内容丰富 ,物体复杂多样 ,但是前景物体和背景图像基本上保持相对独立的特点 ,利用盲信号分离的思想 ,尝试着把独立元分析方法用于视频分割 ,以获得相对独立的背景和前景图像。文中用摄像头采集的连续几帧图像做实验 ,取得了比较满意的效果更多还原

【Abstract】 In this paper, we illustrate the basic background, principle and algorithms of independent component analysis (ICA). Considering the abundance and diversity of the video frequency, which keeps only comparatively independence between the foreground and background, we make use of the idea of blind signal separating, and try the ICA algorithms in video segmentation to get the relatively independent foreground and background. Finally, we tested on a number of continuous frames and got some good results.更多还原