Impulsive noise removal based on sequence multi-scale top-hattransformation

To efficiently remove impulsive noises, a novel algorithm based on sequencemulti-scale top-hat transformation is proposed in this paper. Firstly, themulti-scale white and black top-hat transformations are constructed. Then, thenoise decision functions are introduced to decide and remove the real noisepixels at each scale. Finally, the multi-scale white top-hat transformation,multi-scale black top-hat transformation and noise decision functions areserialized to form the sequence multi-scale top-hat transformation, whichachieves good performance for impulsive noise removal. Experimental results showthat our proposed algorithm is efficient. ICIC International