作者: B. Prados-Suarez , D. Sanchez , J. Chamorro-Martinez
DOI: 10.1109/FUZZY.2008.4630592
关键词:
摘要: In image segmentation it is well known that a given can be analyzed with different detail levels, this why some hierarchical approaches have been proposed to give for each level. Most of these proposals are specially designed precise and defined regions. However regions usually blurred contours, soft color shades, brightness rise the problem imprecision in paper we face both problems considering at definition criteria obtain hierarchy levels. Concretely, propose calculate similarity relation between fuzzy regions, based on two measures take into account transition as likeness their characteristics. Then use nested segmentations by means its alpha-cuts. way tool easily change level new image, just changing value alpha.