Three-dimensional point cloud automatic classification method based on graph convolutional neural network

Abstract

A three-dimensional point cloud classification method based on a graph convolutional neural network is used for automatically classifying and recognizing complex objects in the process that a servicerobot grabs the three-dimensional objects, and it is convenient to determine the grabbing position for grabbing according to category information. The invention belongs to the technical field of three-dimensional perception of computer vision and robot navigation. The method comprises the following steps: (1) preprocessing a three-dimensional point cloud; (2) constructing a point cloud classification network; (3) training the constructed network; and (4) performing a classification test by using the trained network. The method has the advantages that (1) the graph convolutional neural networkis adopted to perform local feature learning, so that the learning ability of the network for local topological structure information is enhanced; (2) global feature learning is carried out on the input point cloud, and the understanding of the network on context semantics is enhanced; the local features and the global features are aggregated, and then the classification score of each point cloudis output by using the full connection layer, so that the method has higher classification precision compared with the existing network.

Description

technical field [0001] In the present invention, a three-dimensional point cloud automatic classification method based on graph convolutional neural network is designed, and by constructing a deep learning network based on graph convolution, automatic classification and recognition of complex objects can be realized in the process of grabbing three-dimensional objects by service robots, which is convenient according to The category information determines where to crawl for crawling. The invention relates to the technical field of three-dimensional perception of computer vision and robot navigation, in particular to a method for automatic classification and recognition of three-dimensional point clouds. Background technique [0002] With the development of theories such as computer vision and artificial intelligence, 3D point clouds are widely used and play an important role in the fields of automotive autonomous driving, robot perception and navigation, and virtual / augmented...

AI Technical Summary

Problems solved by technology

The invention solves the problem of low classification accuracy in the process of three-dimensional point cloud processing, and realizes the function of the service robot to accurately identify the object category in the process of grabbing the object

Images