Title :
Levelings and morphological segmentation
Author_Institution :
Centre de Morphologie Math., Fontainebleau, France
Abstract :
We present a novel class of morphological filters, which are ideal simplification tools in order to facilitate the segmentation of objects of interest. As a subclass of connected operators, they are particularly interesting for segmentation purposes, as they suppress many details in the image while preserving perfectly the position and the sharpness of the structures which are not suppressed. Furthermore, being autodual, they may be applied on the various components of a colour image separately and produce colour image simplifications. We show how to use them in the context of segmentation of sequences, for multimedia applications
Keywords :
filtering theory; image colour analysis; image segmentation; image sequences; multimedia computing; colour image simplifications; connected operators; levelings; morphological filters; morphological segmentation; multimedia applications; object segmentation; sequence segmentation; Filters; Image segmentation;
Conference_Titel :
Computer Graphics, Image Processing, and Vision, 1998. Proceedings. SIBGRAPI '98. International Symposium on
Conference_Location :
Rio de Janeiro
Print_ISBN :
0-8186-9215-4
DOI :
10.1109/SIBGRA.1998.722729