Fast Algorithm for 3D Local Feature Extraction Using Hahn and Charlier Moments

摘要

In this paper, we propose a fast algorithm to extract 3D local features from an object by using Hahn and Charlier moments. These moments have the property to compute local descriptors from a region of interest in an image. This can be realized by varying parameters of Hahn and Charlier polynomials. An algorithm based on matrix multiplication is used to speed up the computational time of 3D moments. The experiment results have illustrated the ability of Hahn and Charlier moments to extract the features from any region of 3D object. However, we have observed the superiority of Hahn moments in terms of reconstruction accuracy. In addition, the proposed algorithm produces a drastic reduction in the computational time as compared with straightforward method.