Implementation of the robust RM-estimators with different influence functions in the RM-KNN filter

In this paper, we present the implementation of the robust RM-estimators with different influence functions such as the cut median and Hampel functions. These functions in the RM algorithms provide the preservation of fine details, impulsive noise removal and multiplicative noise suppression. They demonstrated better robustness in comparison with the simplest cut function. The cut median and Hampel functions were implemented in the RM-KNN filter that is a good tool for preservation of fine details and suppression of noise. The deterministic and statistical properties of the designed filters have been investigated. The optimal parameters of these filters for different noise mixture are presented. The real time implementation by means of use DSP TMS320C6701 demonstrated that the time of processing in the case of use of the simplest cut function is less in comparison with the cut median and Hampel influence functions, but noise suppression is better when cut median or Hampel functions were applied.