Shape reconstruction by analysing all the inner holes

Abstract

Surface reconstruction from point set in 2-dimensional space reconstructs a polygon that best captures the geometric shape of the point set. Input point set can be divided into two types based on the type of sampling used. They are boundary sample and dot pattern. All the existing algorithms in the field of surface reconstruction from point set deals only with one kind of input point set, i.e either with boundary sample or with dot pattern. But ec-shape algorithm is a unified algorithm which work with both types of input point set. Ec-shape algorithm is a delaunay-based method which uses a set of circle and regularity constraints to compute the shape of the input point set. Extensive comparative studies have done with existing methods and it has been observed that ec-shape algorithm perform better than all the existing shape reconstruction algorithm. Major drawback of the algorithm is that it will not detect the inner holes present in the shape. In some cases the shape will be best characterised by the inners holes, eg 8 shape, in such cases a clear output is not produced by the ec shape. The proposed project detect all the inner holes present inside the shape. Holes are characterized through a structural pattern called as fat triangle-arm formed by the Delaunay triangles in the hole regions. The time and space complexities of the algorithm have been shown to be O(n logn) and O(n) respectively, where n is the number of points in S.