Abstract
With the exponential increasement of 3D models, 3D model classification is crucial to the effective management and retrieval of model database. Feature descriptor has important influence on 3D model classification. Voxel descriptor expresses surface and internal information of 3D model. However, it does not contain topological structure information. Shape distribution descriptor expresses geometry relationship of random points on model surface and has rotation invariance. They can all be used to classify 3D models, but accuracy is low due to insufficient description of 3D model. This paper proposes a 3D model classification algorithm that fuses voxel descriptor and shape distribution descriptor. 3D convolutional neural network (CNN) is used to extract voxel features, and 1D CNN is adopted to extract shape distribution features. AdaBoost algorithm is applied to combine several Bayesian classifiers to get a strong classifier for classifying 3D models. Experiments are conducted on ModelNet10, and results show that accuracy of the proposed method is improved.
Highlights
With the development of multimedia technology, 3D model has been applied in many fields such as mechanical design, computer vision, construction industry, entertainment, medical treatment, education, e-commerce, and molecular biology [1]. e number of 3D models is becoming larger and larger in our lives. erefore, the classification of 3D models becomes more and more important
In voxel-based classification method, 3D model is expressed as the distribution of 3D voxel grid. 3D model is denoted by 3D matrix, from which 3D convolutional neural network (CNN) is used to extract features for classifying 3D models [2]
Strong classifier depends on the diversity between weak classifiers and the performance of weak classifiers [7]
Summary
With the development of multimedia technology, 3D model has been applied in many fields such as mechanical design, computer vision, construction industry, entertainment, medical treatment, education, e-commerce, and molecular biology [1]. e number of 3D models is becoming larger and larger in our lives. erefore, the classification of 3D models becomes more and more important. In voxel-based classification method, 3D model is expressed as the distribution of 3D voxel grid. 3D model is denoted by 3D matrix, from which 3D CNN is used to extract features for classifying 3D models [2]. In view-based one, 3D model is converted into a series of 2D images, from which 2D deep learning algorithm is applied to extract view features. En, view features are merged to represent 3D model for model classification [3]. En, CNN is used to extract features from point cloud for classifying 3D models [4]. AdaBoost provides a simple and useful method to generate strong classifier. Shape distribution descriptor is adopted to express geometry relationship of random points on model surface.
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.
