Abstract
In the recent decade, the development of 3D scanners brings the expansion of 3D models, which yields in the increase of demand for developing effective 3D point cloud retrieval methods using only unorganized point clouds instead of mesh data. In this paper, we propose a meshing-free framework for point cloud retrieval by exploiting a bidirectional similarity measurement on local features. Specifically, we first introduce an effective pipeline for keypoint selection by applying principal component analysis to pose normalization and thresholding local similarity of normals. Then, a point cloud based feature descriptor is employed to compute local feature descriptors directly from point clouds. Finally, we propose a bidirectional feature match strategy to handle the similarity measure. Experimental evaluation on a publicly available benchmark demonstrates the effectiveness of our framework and shows it can outperform other alternatives involving state-of-the-art techniques.
Highlights
With the development of 3D data acquisition technologies, point clouds can be generated faster with low cost [1], [2], which leads to rapid growth of the 3D point data stored in databases
We demonstrate in the experiment that our 3D point cloud retrieval framework achieves competitive performance compared against methods based on alternative similarity measures and the state-ofthe-art descriptors
As it has been concluded in Fig. 1, our 3D point cloud retrieval framework includes three main parts: (1) pose normalization, (2) keypoint and feature descriptor extraction, (3) bidirectional feature match, which will be introduced in detail respectively
Summary
With the development of 3D data acquisition technologies, point clouds can be generated faster with low cost [1], [2], which leads to rapid growth of the 3D point data stored in databases. Measuring the similarity between 3D objects is an essential and fundamental task in 3D shape retrieval. A common strategy in shape similarity assessment is to evaluate the similarity score between the shapes in terms of distances with associated feature descriptors. Despite the success of applying similarity measures in mesh retrieval tasks, the evaluation of the effectiveness in 3D point cloud retrieval has been sparsely treated so far. We tackle the problem of 3D point cloud retrieval by extending the bidirectional similarity measure [16] to build a meshing-free framework. We demonstrate in the experiment that our 3D point cloud retrieval framework achieves competitive performance compared against methods based on alternative similarity measures and the state-ofthe-art descriptors. It is important to note that our proposed framework is an efficient, easy-to-apply point cloud retrieval method rather than a learning-based method which is datadependent
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
