A Simple Multi-sensor Mixing Images Separation Algorithm

摘要：

In multi-sensor mixing images, it is a difficult problem to separate super-gaussian and sub-gaussian images. The key is that the probability density function of the image is hard to estimate. To overcome the shortcoming of existing algorithms, in this paper, a parameter-based probability density model is proposed to estimate super-gaussian and sub-gaussian distributions. From the essence of distinguishing super-gaussian and sub-gaussian distributions, mixing images separation is realized by on-line model parameter learning. This model increased separation precision. Simulation shows the proposed algorithm has good performance.

作者: SHI Xiao-fei LI Li

作者单位: Information Engineering College,Dalian Maritime University, Dalian,Liaoning ,PR.China ,116026 Information Engineering College,Dalian University Dalian, Dalian, Liaoning ,PR.China,116622

会议类型: 国际会议

会议名称: 第七届国际测试技术研讨会

会议地点: 北京

会议语种:英文

在线出版日期: 2007-08-05（万方平台首次上网日期，不代表论文的发表时间）