Special Issue Article Open Access

Nonwhite Noise Reduction In Hyperspectral Images

Abstract

Noise reduction is an important preprocessing step to analyze the information in hyperspectral image (HSI).Because the common filtering methods for HSIs is based on the data vectorization or matricization while ignoring the related information between image planes, there are new approaches considering the multidimensional data as whole entities.For example, Multidimensional Wiener filtering (MWF) based on the third order tensor decomposition. To reduce the nonwhite noise from HSIs, the first step is to whiten the noise in HSIs through a prewhitening procedure. Then MWF can help to denoise the prewhitened data.Atlast an inverse prewhitening process can rebuild the estimated signal.

Mrs.Suryatheja.M.B, Gokilamani.R

To read the full article Download Full Article | Visit Full Article