Object recognition based on three-dimensional model

Abstract

It is a challenging work to achieve viewpoint independent object recognition. A new efficient method of object recognition based on 3D model is proposed in this paper. Firstly, we obtain multiple 2D projected images of a single 3D model from different directions, and then extract the normalized Fourier Descriptors of the object's edge in the projected images. According to the fact that 2D projection images within limited view range have continuity and similarity, projections can be clustered into the multiple view feature model, leading to an appropriate number of cluster classes and increases the recognition rate. Finally, the SVM classifier is used for recognition. The experiment results show the effectiveness and efficiency of method proposed. © 2012 Springer-Verlag.