Title : 
Bivariate median filters: new definitions and properties
         
        
            Author : 
Kassam, Saleem A. ; Himayat, Nageen
         
        
            Author_Institution : 
Dept. of Electr. Eng., Pennsylvania Univ., Philadelphia, PA, USA
         
        
        
        
        
            Abstract : 
The authors present some new ideas for defining bivariate median filters, which can in principle also be extended to p-variate data for p> 2. Sensors or channels were used to obtain information about a signal waveform. The authors begin with a brief review of some of the key definitions and properties which have been established so far for bivariate and p-variate filters of the median type. The authors examine one simple definition of the bivariate median which can be useful in defining bivariate filters. This is then extended to define a class of bivariate medians for data points which is based directly on symmetry arguments. Some simulation results and examples illustrating the potential of filters based on these definitions are presented
         
        
            Keywords : 
digital filters; bivariate median filters; p-variate data; p-variate filters; properties; signal waveform; simulation results; Additive noise; Band pass filters; Filtering; Image processing; Narrowband; Radar applications; Radar imaging; Radar signal processing; Signal processing; Sonar;
         
        
        
        
            Conference_Titel : 
Communications, Computers and Signal Processing, 1991., IEEE Pacific Rim Conference on
         
        
            Conference_Location : 
Victoria, BC
         
        
            Print_ISBN : 
0-87942-638-1
         
        
        
            DOI : 
10.1109/PACRIM.1991.160853