Review of ICA Based Fixed-Point Algorithm for Blind Separation of Mixed Images

摘要：

The blind separation of mixed images is a veryexciting area of research. However, classical techniques such aseigen and singular value decomposition, which are based onsecond order statistics, fail to blindly separate mixed signalsin many circumstances. A rapidly developed statistical methodduring last few years, Independent Component Analysis (ICA),which is based on higher order statistics, aims at searching for the components in the mixed signals that are statistically asindependent from each other as possible. This paper introducesthe fundamental theory and basic model of ICA, and analyzes themath principle of frequently-used fast fixed point algorithm forICA, and applies the algorithm in blind separation of randomlymixed images. The results shows that the algorithm is veryeffective and reliable.

作者: Chao Ma Lian-min Wang

作者单位: Faculty of Science Xian Jiaotong University Xian 710049, P.R.China

会议类型: 国际会议

会议名称: The 4th International Conference on Bioinformatics and Biomedical Engineering(第四届IEEE生物信息与生物医学工程国际会议 iCBBE 2010)

会议地点: 成都

会议语种:英文

页码: 1-3

在线出版日期: 2010-06-18（万方平台首次上网日期，不代表论文的发表时间）