3D Point Correspondence by Minimum Description Length in Feature Space

Abstract. Finding point correspondences plays an important role in automatically building statistical shape models from a training set of 3D surfaces. For the point correspondence problem, Davies et al. [4] proposed a minimum-descriptionlength-based objective function to balance the training errors and generalization ability. A recent evaluation study [12] that compares several well-known 3D point correspondence methods for modeling purposes shows that the MDL-based approach [4] is the best method. We adapt the MDL-based objective function for a feature space that can exploit nonlinear properties in point correspondences, and propose an efficient optimization method to minimize the objective function directly in the feature space, given that the inner product of any vector pair can be computed in the feature space. We further employ a Mercer kernel [15] to define the feature space implicitly. A key aspect of our proposed framework is the generalization of the MDL-based objective func...

• Type: Conference
• Year: 2010
• Where: ECCV
• Keywords: Computer Vision, ECCV 2010, Feature Space, Objective Function, Point Correspondences