A multi-scale morphological algorithm for color images segmentation is proposed in this paper.At first

color watershed transform is applied based on tensor gradient to get initial segmentation

i.e.local level connected sets.Both area and color are considered to calculate the similarity between adjacent regions.Then

a RAG(Region Adjacency Graph) and a NNG(Nearest Neighbor Graph) are constructed to accelerate following morphological processing.Finally

color morphological operators are defined based on a special full order relation of color vectors in HSV model.Morphological opening-closing is implemented using vertex-collapse algorithm.A nonlinear scale space will be constructed efficiently with perfect shape preserving.Since extrema in an image are always associated with objects

particular objects are extracted at different scales.Entropy criterion is adopted to control the content contained in the final scale tree.