Abstract: In engineering applications,retrieval of existing 3D CAD models and reuse of their design information can reduce time and cost.This paper describes a freeform surface similarity assessment approach based on region segmentation.A freeform surface is segmented in some regions,which have the similar features of Gaussian curvatures,according to the curvatures of the surface.Thus,a freeform surface is represented with a number of regions.For each region,a 7-dimensional vector is used to describe its geometric properties and topological attributes.Then,a group of the 7-dimensional vectors acts as the shape descriptor of the freeform surfaces.When the similarity between two shape descriptors is assessed,each 7-dimensional vector is regarded as a node with attributes.Two groups of nodes that correspond to the two shape descriptors can be seen as a complete bigraph.The distances between the nodes which belong different groups are considered as the weight of the bigraph.Then,Kuhn-Munkres algorithm is used to seek a matching of maximum total weight from the weighted of the bigraph.The total weight which is the value of the optimal solution is the similar coefficient of the two surfaces.Experiments show that this algorithm can efficiently give the similar degree of the freeform surface models.