Improved 3D Local Feature Descriptor Based on Rotational Projection Statistics and Depth Information

3D local feature descriptor construction is a very challenging task in the field of 3D model analysis. In this paper, an improved Rotational Projection Statistics (IRoPS) descriptor is proposed. For each feature point, the local coordinate system is firstly built and its neighboring points are normalized. Then the normalized neighboring points are rotated and projected onto three coordinate planes. For each rotation, the distribution matrix is computed and the sub-descriptor can be obtained using the central moment, the Shannon entropy, the mean and variance of local depth values. Finally the IRoPS descriptor is constructed by concatenating all the sub-descriptors into a vector. Compared with the Rotational Projection Statistics (RoPS) descriptor, the IRoPS descriptor includes the local depth information and it has better discriminative power. Extensive experiments are performed to verify the superior performance of the proposed descriptor.

[1]  Mark Goadrich,et al.  The relationship between Precision-Recall and ROC curves , 2006, ICML.

[2]  Mohammed Bennamoun,et al.  Rotational Projection Statistics for 3D Local Surface Description and Object Recognition , 2013, International Journal of Computer Vision.

[3]  Tipu Aziz,et al.  Utility of multimaterial 3D printers in creating models with pathological entities to enhance the training experience of neurosurgeons. , 2014, Journal of neurosurgery.

[4]  Mohammed Bennamoun,et al.  A Novel Representation and Feature Matching Algorithm for Automatic Pairwise Registration of Range Images , 2005, International Journal of Computer Vision.

[5]  Ryutarou Ohbuchi,et al.  Shape-similarity search of three-dimensional models using parameterized statistics , 2002, 10th Pacific Conference on Computer Graphics and Applications, 2002. Proceedings..

[6]  Michael G. Strintzis,et al.  This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE Snapshots: A Novel Local Surface , 2022 .

[7]  Anthony Dick,et al.  Special Issue -selected Papers from Dicta 2007 Local 3d Structure Recognition in Range Images , 2022 .

[8]  Ming Ouhyoung,et al.  On Visual Similarity Based 3D Model Retrieval , 2003, Comput. Graph. Forum.

[9]  Andrew E. Johnson,et al.  Using Spin Images for Efficient Object Recognition in Cluttered 3D Scenes , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Marc Levoy,et al.  A volumetric method for building complex models from range images , 1996, SIGGRAPH.

[11]  Chin Seng Chua,et al.  Point Signatures: A New Representation for 3D Object Recognition , 1997, International Journal of Computer Vision.

[12]  Bernt Schiele,et al.  3D object recognition from range images using local feature histograms , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[13]  Federico Tombari,et al.  Unique Signatures of Histograms for Local Surface Description , 2010, ECCV.

[14]  Hui Chen,et al.  3D free-form object recognition in range images using local surface patches , 2004, Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004..

[15]  Mohammed Bennamoun,et al.  On the Repeatability and Quality of Keypoints for Local Feature-based 3D Object Retrieval from Cluttered Scenes , 2009, International Journal of Computer Vision.

[16]  Sven J. Dickinson,et al.  Skeleton based shape matching and retrieval , 2003, 2003 Shape Modeling International..

[17]  Mohammed Bennamoun,et al.  Three-Dimensional Model-Based Object Recognition and Segmentation in Cluttered Scenes , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[18]  Yue Gao,et al.  View-Based Discriminative Probabilistic Modeling for 3D Object Retrieval and Recognition , 2013, IEEE Transactions on Image Processing.