Abstract
When a linear mixture of independent sources is contaminated by multiplicative noise, also called speckle noise, the statistic of the outputs of a linear transformation of the noise data is very different from the statistic that appears when the speckle noise is not present. Specifically, it is not possible find a linear transformation that provides independent outputs and it is necessary study the statistical structure that appears in this case. In this paper, a general approach to obtain the mixture when there exists speckle noise is developed. In order to do this, the linear transformation is searches as the one that reproduces the this theoretical statistic structure. © Springer-Verlag 2004.
Cite
CITATION STYLE
Blanco, D., Mulgrew, B., McLaughlin, S., Ruiz, D. P., & Carrion, M. C. (2004). The use of ICA in speckle noise. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3195, 248–255. https://doi.org/10.1007/978-3-540-30110-3_32
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
