Generalized vector medians for correlated channels
Yinbo Li, Gonzalo R. Arce, J. Bacca · 2004
Inspired by the maximum likelihood (ML) estimates of loca-tion in multivariate spaces, we introduce in this paper a new ltering structure capable of capturing and exploiting both spatial and cross-channel correlations embedded in the data. An adaptive optimization algorithm for a sub-optimal real-ization of the proposed generalized vector median (GVM) lter, namely the marginal GVM, is derived. The effective-ness of the algorithm is shown through a color image denois-ing experiment. 1.