Portal de Periódicos da CAPES

Filtros Robustos RM-KNN con Diferentes Funciones de Influencia para Supresión de Ruido Impulsivo en Imágenes Digitales

2003; National Polytechnic Institute; Volume: 6; Issue: 3 Linguagem: Inglês

10.13053/cys-6-3-1032

2007-9737

Francisco J. Gallegos‐Funes, Volodymyr Ponomaryov, Oleksiy Pogrebnyak, Luis Niño de Rivera,

Advanced Data Compression Techniques

WE PRESENT THE IMPLEMENTATION OF ROBUST IMAGE FILTERING FOR IMPULSIVE NOIVE SUPPRESSION WITH DETAIL PRESERVATION. THE FILTERING SCHEMES USE A SIMILAR TO KNN FILTER TECHNIQUE TO PROVIDE FINE DETAIL PRESRVATION AND THE REDESCENDING M-ESTIMATORS COMBINED WITH THE MEDAIL OR WILCOXON ESTIMATOR TO PROVEDE IMPULSIVE NOISE REJECTION. WE USE DIFFERENT TYPES OF INFLUENCE FUNCTIONS IN THE M-ESTIMATOR TO PROVIDE BETTER IMPULSIVE NOISE REJECTION. THE FILTRATION OF IMAGE SEQUENCES CORRUPTED BY IMPULSIVE NOISE DEMOSTRATES THAT THE PROPOSED METHODS POTENTIALLY COULD PROVIDE A SOLUTION TO QUALITY TV/VIDEO TRASMISSION. THE EFFICIENCY OF THE PROPOSED FILTERS HAS BEEN EVALUATED BY NEMEROUS SIMULATIONS. THE IMPLEMENTATION OF PROPOSED FILTERS WAS IN REAL-TIME BY ME