Image restoration using new spatially-variant morphological filters

In view of the selection of structuring elements problem in morphological filters, this paper presents a new method to generate structuring elements for spatially-variant (SV) morphology. This method takes the theory of amoeba morphology as a foundation and does distance transform from soft boundary to inner hard center in predefined neighborhood through a metric according to the gradient criteria built firstly. The proposed strategy essentially consists in a weighted averaging combining both spatial and tonal information. By the use of the generated structuring elements, SV alternating sequential filter (SVASF) and SV alternating sequential median filter (SVASMF) were established, then, the new filters are compared with spatially-invariant (SI) filters and traditional amoeba filters in noise removing performance. Results on gray-level images show the ability of new SV morphological operators for adaptively preserving the main structures in the image while reducing the noise.