A NEW APPROACH TO VECTOR MEDIAN FILTERING BASED ON SPACE-FILLING CURVES

The availability of a aide set of multidimensional information sources in different application fields (e.g., color cameras, multispectral r emote sensing imagery devices, etc.) is the basis for the interest of image processing research on extensions of scalar nonlinear filtering approaches to multidimensional data filtering. In this paper, a new ap proach to multidimensional median filtering is presented. The method i s structured into two steps, Absolute sorting of the vectorial space b ased on Peano space filling curves is proposed as a preliminary step i n order to map vectorial data onto an appropriate one-dimensional (1-D ) space. Then, a scalar median filtering operation is applied, The mai n advantage of the proposed approach is the computational efficiency o f the absolute sorting step, which makes the method globally faster th an existing median filtering techniques, This is particularly importan t when dealing with a large amount of data (e.g., image sequences), Pr esented results also show that the filtering performances of the propo sed approach are comparable with those of vector median filters presen ted in the literature.