J. Ruiz-Shulcloper and W.G. Kropatsch (Eds.): CIARP 2008, LNCS 5197, pp. 54–61, 2008.
© Springer-Verlag Berlin Heidelberg 2008
Vector Median M-Type L Filter to Process Multichannel
Images
Antonio Toledo-Lopez, Francisco J. Gallegos-Funes, and Volodymyr Ponomaryov
National Polytechnic Institute of Mexico
Mechanical and Electrical Engineering Higher School
UPALM Zacatenco, Av. IPN s/n, Edificio Z, Acceso 3, 3er. piso;
Col. Lindavista, 07738, Mexico D.F., Mexico
Phone/Fax: +52-55-57296000x54622
fgallegosf@ipn.mx
Abstract. The Vector Median M-type L (VMML) -filter to remove impulsive
noise from color images and video color sequences is presented. This filter util-
izes multichannel image processing by using the vector approach and the Me-
dian M-Type L (MML) algorithm. Simulation results indicate that the proposed
filter consistently outperforms other color image filters by balancing the trade-
off between noise suppression, detail preservation, and color retention.
Keywords: Median M-Type L algorithm, Multichannel image processing, Im-
pulsive noise suppression.
1 Introduction
Many useful techniques of multichannel signal processing based on vector processing
have been investigated due to the inherent correlation that exists between the image
channels compared to traditional component-wise approaches [1,2]. Many applica-
tions of this technique are color image processing, remote sensing, robot vision,
biomedical image processing, and high-definition television (HDTV) [1]. Different
filtering techniques have been proposed for color imaging. Particularly, nonlinear
filters applied to color images have been designed to preserve edges and details, and
remove impulsive noise [1-3].
In this paper, we introduce the Vector Median M-Type L (VMML) -filter. This
filter utilizes multichannel image processing by using the vector approach [2,4], and
the Median M-Type L (MML) algorithm [5,6]. The proposed filter uses the combined
RM-estimator [7] into the L-filter [8] by following way. The redescending
M-estimator with different influence functions is combined with the R- (median) es-
timator into the L filtering scheme to obtain sufficient impulsive noise suppression for
each channel by using the vector approach. We also introduce the use of an impulsive
noise detector [9] to improve the properties of noise suppression and detail preserva-
tion in the proposed filtering scheme.
Simulation results have demonstrated that the proposed VMML filter consistently
outperforms other color image filters by balancing the tradeoff between noise sup-
pression, detail preservation, and color retention.