Home > Publications database > Median filtering and its extensions for color images |
Conference Presentation (After Call) | FZJ-2017-07088 |
2017
Abstract: The construction of structure-preserving denoising filters for color images is a challenging task. A new approach is presented that is based on a recently proposed transformation from the RGB color space to the space of symmetric 2x2 matrices (Burgeth, B., Kleefeld A. (2014) An approach to color-morphology based on Einstein addition and Loewner order, Pattern Recognition Letters, 47, 29-39.). This new framework coupled with spatial adaptivity via morphological amoebas offers excellent capabilities for structure-preserving filtering of color images. Additionally, a generalization of the median-based concept is proposed leading to color-valued amoeba M-smoothers. Numerical experiments confirm the applicability and the potential of this novel approach (Kleefeld, A. et al. (2015) Adaptive Filters for Color Images: Median Filtering and its Extensions, Lecture Notes in Computer Science, Springer, Berlin, 9016, 149-158.).
![]() |
The record appears in these collections: |