Three-dimensional model triangular facet feature learning classification method and device based on deep learning

Abstract

The invention provides a three-dimensional model triangular facet feature learning classification method and a three-dimensional model triangular facet feature learning classification device based on deep learning. The three-dimensional model triangular facet feature learning classification method comprises the steps of: constructing a deep convolutional neural network feature learning model; training the deep convolutional neural network feature learning model; performing feature extraction and feature vector construction on a three-dimensional model triangular patch without a category tag, and adopting a Bag-of-words algorithm for reconstructing features in constructed feature vectors; and determining and classifying output features corresponding to the three-dimensional model triangular patch without a category tag according to the trained deep convolutional neural network feature learning model and initial features corresponding to the three-dimensional model triangular patch without a category tag. The method improves the capability of describing the three-dimensional model triangular patch, and ensures the accuracy of feature learning and classification of the three-dimensional model triangular patch.

Description

technical field [0001] The present invention relates to computer technology, in particular to a method and device for learning and classifying features of triangular surfaces of three-dimensional models based on deep learning. Background technique [0002] With the continuous progress and development of social science and technology, 3D technology has become an important part of modern technology. As one of the important basic technologies for 3D model understanding and processing, 3D model triangle surface feature learning and classification technology plays a huge role in various 3D technology fields such as 3D modeling, 3D animation, and 3D mapping. [0003] In the prior art, a variety of 3D model triangle surface feature learning and classification techniques have been proposed. For example, in 2016, Zhenyu Shu et al. of Zhejiang University proposed an unsupervised deep learning-based 3D model triangle surface classification and co-segmentation method. . Based on pre-s...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a three-dimensional model triangular surface feature learning and classification method and de...

Images